{"id":37,"date":"2018-05-03T15:21:56","date_gmt":"2018-05-03T13:21:56","guid":{"rendered":"http:\/\/michaelkamp.org\/mk_v1\/?page_id=37"},"modified":"2023-12-21T16:51:45","modified_gmt":"2023-12-21T15:51:45","slug":"publications","status":"publish","type":"page","link":"https:\/\/michaelkamp.org\/?page_id=37","title":{"rendered":"Publications"},"content":{"rendered":"<div class=\"teachpress_pub_list\"><form name=\"tppublistform\" method=\"get\"><a name=\"tppubs\" id=\"tppubs\"><\/a><div class=\"teachpress_cloud\"><span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=18&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">black-box<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=41&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">black-box parallelization<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=60&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">causality<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=76&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">CKD<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=25&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">communication-efficient<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=15&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">corresponding projections<\/a><\/span> <span style=\"font-size:35px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"10 Publications\" class=\"\">deep learning<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=13&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">DiSIEM<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=7&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">distributed<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=24&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">dynamic averaging<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=19&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">federated<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=61&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">federated causal discovery<\/a><\/span> <span style=\"font-size:28px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"8 Publications\" class=\"\">federated learning<\/a><\/span> <span style=\"font-size:14px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=29&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"4 Publications\" class=\"\">flatness<\/a><\/span> <span style=\"font-size:14px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=31&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"4 Publications\" class=\"\">generalization<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=77&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">healthcare<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=63&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">interpretable<\/a><\/span> <span style=\"font-size:14px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=20&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"4 Publications\" class=\"\">learning theory<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=28&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">loss surface<\/a><\/span> <span style=\"font-size:21px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"6 Publications\" class=\"\">machine learning<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=78&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">medicine<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=27&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">neural networks<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=4&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">parallelization<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=50&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">relative flatness<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=30&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">robustness<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=49&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">theory of deep learning<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=3&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">unsupervised<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=14&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">visual analytics<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=11&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"3 Publications\" class=\"\">visualization<\/a><\/span> <span style=\"font-size:11px;\"><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=65&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\" title=\"2 Publications\" class=\"\">XAI<\/a><\/span> <\/div><div class=\"teachpress_filter\"><select class=\"default\" name=\"yr\" id=\"yr\" tabindex=\"2\" onchange=\"teachpress_jumpMenu('parent',this, 'https:\/\/michaelkamp.org\/?page_id=37&amp;')\">\r\n                   <option value=\"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=#tppubs\">All years<\/option>\r\n                   <option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2026#tppubs\" >2026<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2025#tppubs\" >2025<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2024#tppubs\" >2024<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2023#tppubs\" >2023<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2022#tppubs\" >2022<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2021#tppubs\" >2021<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2020#tppubs\" >2020<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2019#tppubs\" >2019<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2018#tppubs\" >2018<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2017#tppubs\" >2017<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2016#tppubs\" >2016<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2015#tppubs\" >2015<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2014#tppubs\" >2014<\/option><option value = \"tgid=&amp;type=&amp;auth=&amp;usr=&amp;yr=2013#tppubs\" >2013<\/option>\r\n                <\/select><select class=\"default\" name=\"type\" id=\"type\" tabindex=\"3\" onchange=\"teachpress_jumpMenu('parent',this, 'https:\/\/michaelkamp.org\/?page_id=37&amp;')\">\r\n                   <option value=\"tgid=&amp;yr=&amp;auth=&amp;usr=&amp;type=#tppubs\">All types<\/option>\r\n                   <option value = \"tgid=&amp;yr=&amp;auth=&amp;usr=&amp;type=article#tppubs\" >Journal Articles<\/option><option value = \"tgid=&amp;yr=&amp;auth=&amp;usr=&amp;type=inproceedings#tppubs\" >Proceedings Articles<\/option><option value = \"tgid=&amp;yr=&amp;auth=&amp;usr=&amp;type=phdthesis#tppubs\" >PhD Theses<\/option><option value = \"tgid=&amp;yr=&amp;auth=&amp;usr=&amp;type=workshop#tppubs\" >Workshops<\/option>\r\n                <\/select><select class=\"default\" name=\"auth\" id=\"auth\" tabindex=\"5\" onchange=\"teachpress_jumpMenu('parent',this, 'https:\/\/michaelkamp.org\/?page_id=37&amp;')\">\r\n                   <option value=\"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=#tppubs\">All authors<\/option>\r\n                   <option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=107#tppubs\" > Abourayya, Amr<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=21#tppubs\" > Adilova, Linara<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=93#tppubs\" > Ahmadi, Seyed-Ahmad<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=321#tppubs\" > Aizenberg, Michele R<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=89#tppubs\" > Alves, Victor<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=304#tppubs\" > Ambigapathy, Narmada<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=326#tppubs\" > Andrearczyk, Vincent<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=30#tppubs\" > Andrienko, Gennady<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=29#tppubs\" > Andrienko, Natalia<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=109#tppubs\" > Andriushchenko, Maksym<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=353#tppubs\" > Angeles-Valdez, Diego<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=124#tppubs\" > Aqeel, Iram<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=369#tppubs\" > Aresta, Guilherme<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=380#tppubs\" > Audenaert, Emmanuel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=115#tppubs\" > Ayday, Erman<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=387#tppubs\" > Badeli, Vahid<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=397#tppubs\" > Bahnsen, Fin H<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=352#tppubs\" > Balducci, Thania<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=392#tppubs\" > Balzer, Miriam<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=424#tppubs\" > Basu, Oliver<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=349#tppubs\" > Beier, Susann<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=338#tppubs\" > Ben-Hamadou, Achraf<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=32#tppubs\" > Bessani, Alysson<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=113#tppubs\" > Bodic, Pierre Le<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=342#tppubs\" > Bolelli, Federico<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=7#tppubs\" > Boley, Mario<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=74#tppubs\" > Bothe, Sebastian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=341#tppubs\" > Boyer, Edmond<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=406#tppubs\" > Brehmer, Alexander<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=434#tppubs\" > Briq, Rania<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=371#tppubs\" > Campilho, Aur\u00e9lio<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=428#tppubs\" > Chang, Ti-chiun<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=359#tppubs\" > Chatterjee, Soumick<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=390#tppubs\" > Chen, Jianxu<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=28#tppubs\" > Chen, Siming<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=314#tppubs\" > Chen, Xiaojun<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=298#tppubs\" > Chen, Xiaoxi<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=318#tppubs\" > Christ, Patrick Ferdinand<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=436#tppubs\" > Cohen, Sarel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=336#tppubs\" > Cornelissen, Stefan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=370#tppubs\" > Cunha, Ant\u00f3nio<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=108#tppubs\" > Dada, Amin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=118#tppubs\" > Dalleiger, Sebastian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=80#tppubs\" > D\u00e4ubener, Sina<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=3#tppubs\" > Deligiannakis, Antonios<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=327#tppubs\" > Depeursinge, Adrien<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=332#tppubs\" > Dorent, Reuben<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=55#tppubs\" > Dou, Qi<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=360#tppubs\" > Dubost, Florian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=410#tppubs\" > Dudda, Marcel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=421#tppubs\" > Duong, Luc<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=377#tppubs\" > Duquesne, Kate<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=313#tppubs\" > Eagleson, Roy<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=94#tppubs\" > Egger, Jan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=415#tppubs\" > Elhabian, Shireen<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=320#tppubs\" > Ellis, David G<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=358#tppubs\" > Ferrari, Vincenzo<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=87#tppubs\" > Ferreira, Andr\u00e9<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=368#tppubs\" > Ferreira, Carlos<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=412#tppubs\" > Fink, Matthias A<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=81#tppubs\" > Fischer, Asja<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=83#tppubs\" > Fischer, Jonas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=110#tppubs\" > Fischer, Michael Kamp Asja<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=125#tppubs\" > Fitzpatrick, Tim<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=1#tppubs\" > Flouris, Ioannis<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=393#tppubs\" > Fragemann, Jana<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=356#tppubs\" > Frayne, Richard<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=435#tppubs\" > Fried, Ohad<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=301#tppubs\" > Friedrich, Paul<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=373#tppubs\" > Garcia, Jose<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=4#tppubs\" > Garofalakis, Minos<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=8#tppubs\" > G\u00e4rtner, Thomas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=351#tppubs\" > Garza-Villarreal, Eduardo A<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=322#tppubs\" > Gatidis, Sergios<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=348#tppubs\" > Gharleghi, Ramtin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=2#tppubs\" > Giatrakos, Nikos<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=35#tppubs\" > Giesselbach, Sven<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=127#tppubs\" > Gou, Liang<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=343#tppubs\" > Grana, Costantino<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=294#tppubs\" > Gsaxner, Christina<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=37#tppubs\" > Gunar Ernis, Michael Kamp<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=389#tppubs\" > Gunzer, Matthias<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=364#tppubs\" > Haehn, Daniel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=432#tppubs\" > Han, Ting<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=408#tppubs\" > Hanusrichter, Yannik<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=328#tppubs\" > Hatt, Mathieu<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=384#tppubs\" > Heidemeyer, Hauke<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=407#tppubs\" > Heine, Lukas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=52#tppubs\" > Heinrich, Danny<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=325#tppubs\" > Heller, Nicholas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=50#tppubs\" > Heppe, Lukas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=123#tppubs\" > Hidalgo, Guillermo<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=131#tppubs\" > Hofmann, Thomas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=386#tppubs\" > H\u00f6lzle, Frank<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=405#tppubs\" > H\u00f6rst, Fabian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=99#tppubs\" > Hou, Yijie<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=423#tppubs\" > Hoyer, Peter F<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=23#tppubs\" > H\u00fcger, Fabian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=401#tppubs\" > Jaeger, Paul F<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=111#tppubs\" > Jaggi, Martin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=399#tppubs\" > Jaus, Alexander<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=357#tppubs\" > Ji, Yuanfeng<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=58#tppubs\" > Jiang, Meirui<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=303#tppubs\" > Jin, Yuan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=365#tppubs\" > John, Christoph<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=134#tppubs\" > Jonske, Enrico Nasca Moon Kim Frederic<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=394#tppubs\" > Jonske, Frederic<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=103#tppubs\" > Kaltenpoth, David<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=5#tppubs\" > Kamp, Michael<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=10#tppubs\" > Keren, Daniel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=437#tppubs\" > Kesselheim, Stefan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=413#tppubs\" > Keyl, Julius<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=104#tppubs\" > Kim, Moon<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=90#tppubs\" > Kim, Moon-Sung<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=316#tppubs\" > Kirschke, Jan Stefan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=92#tppubs\" > Kleesiek, Jens<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=13#tppubs\" > Kopp, Christine<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=381#tppubs\" > Krebs, Claudia<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=388#tppubs\" > Krieger, Kathrin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=126#tppubs\" > Krishnan, Sriram<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=335#tppubs\" > Kujawa, Aaron<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=323#tppubs\" > K\u00fcstner, Thomas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=374#tppubs\" > Lalande, Alain<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=416#tppubs\" > Lamecker, Hans<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=337#tppubs\" > Langenhuizen, Patrick<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=382#tppubs\" >van Leeuwen, Timo<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=319#tppubs\" > Li, Hongwei Bran<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=86#tppubs\" > Li, Jianning<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=299#tppubs\" > Li, Wenxuan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=57#tppubs\" > Li, Xiaoxiao<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=96#tppubs\" > Li, Yun<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=331#tppubs\" > Liebl, Hans<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=106#tppubs\" >Michael Kamp Linara Adilova, Gennady Andrienko<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=404#tppubs\" > Lindo, Mariana<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=79#tppubs\" > Linsner, Florian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=330#tppubs\" > L\u00f6ffler, Maximilian T<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=295#tppubs\" > Luijten, Gijs<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=344#tppubs\" > Lumetti, Luca<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=429#tppubs\" > Luo, Ping<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=346#tppubs\" > Ma, Jun<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=425#tppubs\" > Maal, Thomas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=315#tppubs\" > M\u00e4chler, Heinrich<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=396#tppubs\" > Malorodov, Stanislav<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=15#tppubs\" > Manea, Andrei<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=400#tppubs\" > Marinov, Zdravko<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=121#tppubs\" > Masud, Avais<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=362#tppubs\" > Mattern, Hendrik<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=14#tppubs\" > May, Michael<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=391#tppubs\" >van Meegdenburg, Timo<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=378#tppubs\" > Mekhzoum, Hamza<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=307#tppubs\" > Melito, Gian Marco<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=309#tppubs\" > Memon, Afaque R<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=430#tppubs\" > Menze, Bjoern<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=84#tppubs\" > Mian, Osman<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=16#tppubs\" > Missura, Olana<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=6#tppubs\" > Mock, Michael<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=53#tppubs\" > Morik, Katharina<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=305#tppubs\" > Nasca, Enrico<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=41#tppubs\" > Natious, Livin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=91#tppubs\" > Nensa, Felix<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=27#tppubs\" > Nguyen, Phong H.<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=366#tppubs\" > N\u00fcrnberger, Andreas<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=376#tppubs\" > Oevelen, Aline Van<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=85#tppubs\" >David Kaltenpoth Osman Mian, Michael Kamp<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=418#tppubs\" > Paniagua, Beatriz<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=312#tppubs\" > Patel, Rajnikant<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=36#tppubs\" > Paurat, Daniel<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=367#tppubs\" > Pedrosa, Jo\u00e3o<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=293#tppubs\" > Pepe, Antonio<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=39#tppubs\" > Petzka, Henning<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=51#tppubs\" > Piatkowski, Nico<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=411#tppubs\" > Podleska, Lars E<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=340#tppubs\" > Pujades, Sergi<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=88#tppubs\" > Puladi, Behrus<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=296#tppubs\" > Qu, Chongyu<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=116#tppubs\" > Rao, Bharat<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=114#tppubs\" > Rao, Kanishka<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=339#tppubs\" > Rekik, Ahmed<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=395#tppubs\" > Rempe, Moritz<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=431#tppubs\" > Reyes, Mauricio<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=311#tppubs\" > Ribaupierre, Sandrine De<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=355#tppubs\" > Rittner, Leticia<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=385#tppubs\" > R\u00f6hrig, Rainer<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=317#tppubs\" > Rosa, Ezequiel De La<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=42#tppubs\" > Rosenzweig, Julia<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=119#tppubs\" > Salazer, Thomas L<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=345#tppubs\" > Salehi, Hamidreza<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=403#tppubs\" > Santos, Ana Sofia<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=427#tppubs\" > Schiele, Gregor<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=310#tppubs\" > Schlachta, Christopher<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=24#tppubs\" > Schlicht, Peter<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=130#tppubs\" > Sch\u00f6lkopf, Bernhard<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=361#tppubs\" > Schreiber, Stefanie<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=11#tppubs\" > Schuster, Assaf<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=398#tppubs\" > Seibold, Constantin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=48#tppubs\" > Seiffarth, Florian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=329#tppubs\" > Sekuboyina, Anjany<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=122#tppubs\" > Serur, David<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=334#tppubs\" > Shapey, Jonathan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=12#tppubs\" > Sharfman, Izchak<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=120#tppubs\" > Sheth, Naitik<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=324#tppubs\" > Shusharina, Nadya<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=22#tppubs\" > Sicking, Joachim<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=129#tppubs\" > Singh, Sidak Pal<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=40#tppubs\" > Sminchisescu, Cristian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=306#tppubs\" > Solak, Naida<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=354#tppubs\" > Souza, Roberto<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=350#tppubs\" > Sowmya, Arcot<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=363#tppubs\" > Speck, Oliver<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=402#tppubs\" > Stiefelhagen, Rainer<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=128#tppubs\" > Sunr, Dong<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=372#tppubs\" > Suter, Yannick<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=31#tppubs\" > Thonnard, Olivier<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=414#tppubs\" > Tserpes, Konstantinos<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=33#tppubs\" > Turkay, Cagatay<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=17#tppubs\" > Ullrich, Katrin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=420#tppubs\" > Urschler, Martin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=379#tppubs\" > Vandemeulebroucke, Jef<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=375#tppubs\" > Vandenbossche, Vicky<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=333#tppubs\" > Vercauteren, Tom<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=383#tppubs\" > Vereecke, Evie<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=18#tppubs\" > Vogt, Martin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=78#tppubs\" > Vreeken, Jilles<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=308#tppubs\" > Vu, Viet Duc<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=419#tppubs\" > Wachinger, Christian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=433#tppubs\" > Walter, Nils Philipp<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=98#tppubs\" > Wang, Chengshun<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=95#tppubs\" > Wang, Junhong<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=422#tppubs\" > Wasserthal, Jakob<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=117#tppubs\" > Webb, Geoffrey I.<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=47#tppubs\" > Welke, Pascal<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=409#tppubs\" > We\u00dfling, Martin<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=25#tppubs\" > Wirtz, Tim<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=426#tppubs\" > Witjes, Max JH<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=300#tppubs\" > Wodzinski, Marek<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=19#tppubs\" > Wrobel, Stefan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=302#tppubs\" > Xie, Kangxian<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=101#tppubs\" > Xue, Xiangyang<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=112#tppubs\" > Yang, Fan<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=292#tppubs\" > Yang, Jiancheng<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=132#tppubs\" > Ye, Erasmo Purificato Jielin Feng Xinwu<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=100#tppubs\" > Zhang, Li<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=297#tppubs\" > Zhang, Tiezheng<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=59#tppubs\" > Zhang, Xiaofei<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=102#tppubs\" > Zhang, Xiaolong<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=347#tppubs\" > Zhang, Yao<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=97#tppubs\" > Zhou, Zhaoyu<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=291#tppubs\" > Zhou, Zongwei<\/option><option value = \"tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=417#tppubs\" > Zuki\u0107, D\u017eenan<\/option>\r\n                <\/select><\/div><\/form><div class=\"tablenav\"><div class=\"tablenav-pages\"><span class=\"displaying-num\">52 entries<\/span> <a class=\"page-numbers button disabled\">&laquo;<\/a> <a class=\"page-numbers button disabled\">&lsaquo;<\/a> 1 of 2 <a href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;limit=2&amp;tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=&amp;tsr=#tppubs\" title=\"next page\" class=\"page-numbers button\">&rsaquo;<\/a> <a href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;limit=2&amp;tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=&amp;tsr=#tppubs\" title=\"last page\" class=\"page-numbers button\">&raquo;<\/a> <\/div><\/div><div class=\"teachpress_publication_list\"><h3 class=\"tp_h3\" id=\"tp_h3_2026\">2026<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Mian, Osman;  Kleesiek, Jens;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a class=\"tp_title_link\" onclick=\"teachpress_pub_showhide('51','tp_links')\" style=\"cursor:pointer;\">Unified Causal Discovery and Missing Data Imputation<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">The 29th International Conference on Artificial Intelligence and Statistics, <\/span><span class=\"tp_pub_additional_year\">2026<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_51\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('51','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_51\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('51','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_51\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{mianunified,<br \/>\r\ntitle = {Unified Causal Discovery and Missing Data Imputation},<br \/>\r\nauthor = {Osman Mian and Jens Kleesiek and Michael Kamp},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2026\/05\/958_Unified_Causal_Discovery_a.pdf},<br \/>\r\nyear  = {2026},<br \/>\r\ndate = {2026-01-01},<br \/>\r\nurldate = {2026-01-01},<br \/>\r\nbooktitle = {The 29th International Conference on Artificial Intelligence and Statistics},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('51','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_51\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2026\/05\/958_Unified_Causal_Discovery_a.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2026\/05\/958_Unified_Causal_Discovery_[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2026\/05\/958_Unified_Causal_Discovery_[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('51','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Walter, Nils Philipp;  Adilova, Linara;  Vreeken, Jilles;  Kamp, Michael<\/p><p class=\"tp_pub_title\">When flatness does (not) guarantee adversarial robustness <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">International Conference on Learning Representations, <\/span><span class=\"tp_pub_additional_year\">2026<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_52\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('52','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_52\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{walter2025flatness,<br \/>\r\ntitle = {When flatness does (not) guarantee adversarial robustness},<br \/>\r\nauthor = {Nils Philipp Walter and Linara Adilova and Jilles Vreeken and Michael Kamp},<br \/>\r\nyear  = {2026},<br \/>\r\ndate = {2026-01-01},<br \/>\r\nbooktitle = {International Conference on Learning Representations},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('52','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Briq, Rania;  Kamp, Michael;  Fried, Ohad;  Cohen, Sarel;  Kesselheim, Stefan<\/p><p class=\"tp_pub_title\">Exploring and Exploiting Stability in Latent Flow Matching <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">International Conference on Machine Learning, <\/span><span class=\"tp_pub_additional_year\">2026<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_55\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('55','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_55\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{briq2026exploring,<br \/>\r\ntitle = {Exploring and Exploiting Stability in Latent Flow Matching},<br \/>\r\nauthor = {Rania Briq and Michael Kamp and Ohad Fried and Sarel Cohen and Stefan Kesselheim},<br \/>\r\nyear  = {2026},<br \/>\r\ndate = {2026-01-01},<br \/>\r\nbooktitle = {International Conference on Machine Learning},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('55','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2025\">2025<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Abourayya, Amr;  Kleesiek, Jens;  Rao, Kanishka;  Ayday, Erman;  Rao, Bharat;  Webb, Geoffrey I.;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=873\">Little is Enough: Boosting Privacy by Sharing Only Hard Labels in Federated Semi-Supervised Learning<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), <\/span><span class=\"tp_pub_additional_publisher\">AAAI, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_42\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('42','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=74#tppubs\" title=\"Show all publications which have a relationship to this tag\">aimhi<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=75#tppubs\" title=\"Show all publications which have a relationship to this tag\">FedCT<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=38#tppubs\" title=\"Show all publications which have a relationship to this tag\">semi-supervised<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_42\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{abourayya2025little,<br \/>\r\ntitle = {Little is Enough: Boosting Privacy by Sharing Only Hard Labels in Federated Semi-Supervised Learning},<br \/>\r\nauthor = {Amr Abourayya and Jens Kleesiek and Kanishka Rao and Erman Ayday and Bharat Rao and Geoffrey I. Webb and Michael Kamp},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-02-27},<br \/>\r\nurldate = {2025-02-27},<br \/>\r\nbooktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},<br \/>\r\npublisher = {AAAI},<br \/>\r\nkeywords = {aimhi, FedCT, federated learning, semi-supervised},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('42','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Dalleiger, Sebastian;  Vreeken, Jilles;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=874\">Federated Binary Matrix Factorization using Proximal Optimization<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), <\/span><span class=\"tp_pub_additional_publisher\">AAAI, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_43\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('43','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_43\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{dalleiger2025federated,<br \/>\r\ntitle = {Federated Binary Matrix Factorization using Proximal Optimization},<br \/>\r\nauthor = {Sebastian Dalleiger and Jilles Vreeken and Michael Kamp},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-02-27},<br \/>\r\nurldate = {2025-02-27},<br \/>\r\nbooktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},<br \/>\r\npublisher = {AAAI},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('43','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Federated Binary Matrix Factorization using Proximal Optimization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Dalleiger2023Federated-1.png\" width=\"160\" alt=\"Federated Binary Matrix Factorization using Proximal Optimization\" \/><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Ye, Erasmo Purificato Jielin Feng Xinwu<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=876\">GNNFairViz: Visual Analysis for Graph Neural Network Fairness<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">IEEE Transactions on Visualization and Computer Graphics, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_47\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('47','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_47\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('47','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=80#tppubs\" title=\"Show all publications which have a relationship to this tag\">fairness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=81#tppubs\" title=\"Show all publications which have a relationship to this tag\">GNN<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=82#tppubs\" title=\"Show all publications which have a relationship to this tag\">graph neural networks<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=83#tppubs\" title=\"Show all publications which have a relationship to this tag\">graphs<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=14#tppubs\" title=\"Show all publications which have a relationship to this tag\">visual analytics<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=11#tppubs\" title=\"Show all publications which have a relationship to this tag\">visualization<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_47\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{ye2025gnnfairviz,<br \/>\r\ntitle = {GNNFairViz: Visual Analysis for Graph Neural Network Fairness},<br \/>\r\nauthor = {Erasmo Purificato Jielin Feng Xinwu Ye},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/GNNFairViz_Visual_Analysis_for_Graph_Neural_Network_Fairness.pdf},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-02-17},<br \/>\r\nurldate = {2025-02-17},<br \/>\r\njournal = {IEEE Transactions on Visualization and Computer Graphics},<br \/>\r\npublisher = {IEEE},<br \/>\r\nkeywords = {fairness, GNN, graph neural networks, graphs, visual analytics, visualization},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('47','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_47\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/GNNFairViz_Visual_Analysis_for_Graph_Neural_Network_Fairness.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/GNNFairViz_Visual_Analysis_fo[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/GNNFairViz_Visual_Analysis_fo[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('47','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Li, Jianning;  Zhou, Zongwei;  Yang, Jiancheng;  Pepe, Antonio;  Gsaxner, Christina;  Luijten, Gijs;  Qu, Chongyu;  Zhang, Tiezheng;  Chen, Xiaoxi;  Li, Wenxuan;  Wodzinski, Marek;  Friedrich, Paul;  Xie, Kangxian;  Jin, Yuan;  Ambigapathy, Narmada;  Nasca, Enrico;  Solak, Naida;  Melito, Gian Marco;  Vu, Viet Duc;  Memon, Afaque R;  Schlachta, Christopher;  Ribaupierre, Sandrine De;  Patel, Rajnikant;  Eagleson, Roy;  Chen, Xiaojun;  M\u00e4chler, Heinrich;  Kirschke, Jan Stefan;  Rosa, Ezequiel De La;  Christ, Patrick Ferdinand;  Li, Hongwei Bran;  Ellis, David G;  Aizenberg, Michele R;  Gatidis, Sergios;  K\u00fcstner, Thomas;  Shusharina, Nadya;  Heller, Nicholas;  Andrearczyk, Vincent;  Depeursinge, Adrien;  Hatt, Mathieu;  Sekuboyina, Anjany;  L\u00f6ffler, Maximilian T;  Liebl, Hans;  Dorent, Reuben;  Vercauteren, Tom;  Shapey, Jonathan;  Kujawa, Aaron;  Cornelissen, Stefan;  Langenhuizen, Patrick;  Ben-Hamadou, Achraf;  Rekik, Ahmed;  Pujades, Sergi;  Boyer, Edmond;  Bolelli, Federico;  Grana, Costantino;  Lumetti, Luca;  Salehi, Hamidreza;  Ma, Jun;  Zhang, Yao;  Gharleghi, Ramtin;  Beier, Susann;  Sowmya, Arcot;  Garza-Villarreal, Eduardo A;  Balducci, Thania;  Angeles-Valdez, Diego;  Souza, Roberto;  Rittner, Leticia;  Frayne, Richard;  Ji, Yuanfeng;  Ferrari, Vincenzo;  Chatterjee, Soumick;  Dubost, Florian;  Schreiber, Stefanie;  Mattern, Hendrik;  Speck, Oliver;  Haehn, Daniel;  John, Christoph;  N\u00fcrnberger, Andreas;  Pedrosa, Jo\u00e3o;  Ferreira, Carlos;  Aresta, Guilherme;  Cunha, Ant\u00f3nio;  Campilho, Aur\u00e9lio;  Suter, Yannick;  Garcia, Jose;  Lalande, Alain;  Vandenbossche, Vicky;  Oevelen, Aline Van;  Duquesne, Kate;  Mekhzoum, Hamza;  Vandemeulebroucke, Jef;  Audenaert, Emmanuel;  Krebs, Claudia; van Leeuwen, Timo;  Vereecke, Evie;  Heidemeyer, Hauke;  R\u00f6hrig, Rainer;  H\u00f6lzle, Frank;  Badeli, Vahid;  Krieger, Kathrin;  Gunzer, Matthias;  Chen, Jianxu; van Meegdenburg, Timo;  Dada, Amin;  Balzer, Miriam;  Fragemann, Jana;  Jonske, Frederic;  Rempe, Moritz;  Malorodov, Stanislav;  Bahnsen, Fin H;  Seibold, Constantin;  Jaus, Alexander;  Marinov, Zdravko;  Jaeger, Paul F;  Stiefelhagen, Rainer;  Santos, Ana Sofia;  Lindo, Mariana;  Ferreira, Andr\u00e9;  Alves, Victor;  Kamp, Michael;  Abourayya, Amr;  Nensa, Felix;  H\u00f6rst, Fabian;  Brehmer, Alexander;  Heine, Lukas;  Hanusrichter, Yannik;  We\u00dfling, Martin;  Dudda, Marcel;  Podleska, Lars E;  Fink, Matthias A;  Keyl, Julius;  Tserpes, Konstantinos;  Kim, Moon-Sung;  Elhabian, Shireen;  Lamecker, Hans;  Zuki\u0107, D\u017eenan;  Paniagua, Beatriz;  Wachinger, Christian;  Urschler, Martin;  Duong, Luc;  Wasserthal, Jakob;  Hoyer, Peter F;  Basu, Oliver;  Maal, Thomas;  Witjes, Max JH;  Schiele, Gregor;  Chang, Ti-chiun;  Ahmadi, Seyed-Ahmad;  Luo, Ping;  Menze, Bjoern;  Reyes, Mauricio<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=878\">Medshapenet\u2013a large-scale dataset of 3d medical shapes for computer vision<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">Biomedical Engineering\/Biomedizinische Technik, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_48\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('48','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_48\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('48','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=84#tppubs\" title=\"Show all publications which have a relationship to this tag\">dataset<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=78#tppubs\" title=\"Show all publications which have a relationship to this tag\">medicine<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=85#tppubs\" title=\"Show all publications which have a relationship to this tag\">shapes<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_48\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{li2025medshapenet,<br \/>\r\ntitle = {Medshapenet\u2013a large-scale dataset of 3d medical shapes for computer vision},<br \/>\r\nauthor = {Jianning Li and Zongwei Zhou and Jiancheng Yang and Antonio Pepe and Christina Gsaxner and Gijs Luijten and Chongyu Qu and Tiezheng Zhang and Xiaoxi Chen and Wenxuan Li and Marek Wodzinski and Paul Friedrich and Kangxian Xie and Yuan Jin and Narmada Ambigapathy and Enrico Nasca and Naida Solak and Gian Marco Melito and Viet Duc Vu and Afaque R Memon and Christopher Schlachta and Sandrine De Ribaupierre and Rajnikant Patel and Roy Eagleson and Xiaojun Chen and Heinrich M\u00e4chler and Jan Stefan Kirschke and Ezequiel De La Rosa and Patrick Ferdinand Christ and Hongwei Bran Li and David G Ellis and Michele R Aizenberg and Sergios Gatidis and Thomas K\u00fcstner and Nadya Shusharina and Nicholas Heller and Vincent Andrearczyk and Adrien Depeursinge and Mathieu Hatt and Anjany Sekuboyina and Maximilian T L\u00f6ffler and Hans Liebl and Reuben Dorent and Tom Vercauteren and Jonathan Shapey and Aaron Kujawa and Stefan Cornelissen and Patrick Langenhuizen and Achraf Ben-Hamadou and Ahmed Rekik and Sergi Pujades and Edmond Boyer and Federico Bolelli and Costantino Grana and Luca Lumetti and Hamidreza Salehi and Jun Ma and Yao Zhang and Ramtin Gharleghi and Susann Beier and Arcot Sowmya and Eduardo A Garza-Villarreal and Thania Balducci and Diego Angeles-Valdez and Roberto Souza and Leticia Rittner and Richard Frayne and Yuanfeng Ji and Vincenzo Ferrari and Soumick Chatterjee and Florian Dubost and Stefanie Schreiber and Hendrik Mattern and Oliver Speck and Daniel Haehn and Christoph John and Andreas N\u00fcrnberger and Jo\u00e3o Pedrosa and Carlos Ferreira and Guilherme Aresta and Ant\u00f3nio Cunha and Aur\u00e9lio Campilho and Yannick Suter and Jose Garcia and Alain Lalande and Vicky Vandenbossche and Aline Van Oevelen and Kate Duquesne and Hamza Mekhzoum and Jef Vandemeulebroucke and Emmanuel Audenaert and Claudia Krebs and Timo van Leeuwen and Evie Vereecke and Hauke Heidemeyer and Rainer R\u00f6hrig and Frank H\u00f6lzle and Vahid Badeli and Kathrin Krieger and Matthias Gunzer and Jianxu Chen and Timo van Meegdenburg and Amin Dada and Miriam Balzer and Jana Fragemann and Frederic Jonske and Moritz Rempe and Stanislav Malorodov and Fin H Bahnsen and Constantin Seibold and Alexander Jaus and Zdravko Marinov and Paul F Jaeger and Rainer Stiefelhagen and Ana Sofia Santos and Mariana Lindo and Andr\u00e9 Ferreira and Victor Alves and Michael Kamp and Amr Abourayya and Felix Nensa and Fabian H\u00f6rst and Alexander Brehmer and Lukas Heine and Yannik Hanusrichter and Martin We\u00dfling and Marcel Dudda and Lars E Podleska and Matthias A Fink and Julius Keyl and Konstantinos Tserpes and Moon-Sung Kim and Shireen Elhabian and Hans Lamecker and D\u017eenan Zuki\u0107 and Beatriz Paniagua and Christian Wachinger and Martin Urschler and Luc Duong and Jakob Wasserthal and Peter F Hoyer and Oliver Basu and Thomas Maal and Max JH Witjes and Gregor Schiele and Ti-chiun Chang and Seyed-Ahmad Ahmadi and Ping Luo and Bjoern Menze and Mauricio Reyes},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/li_medhsapenet.pdf},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-02-01},<br \/>\r\nurldate = {2025-02-01},<br \/>\r\njournal = {Biomedical Engineering\/Biomedizinische Technik},<br \/>\r\npublisher = {De Gruyter},<br \/>\r\nkeywords = {dataset, medicine, shapes},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('48','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_48\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/li_medhsapenet.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/li_medhsapenet.pdf\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/li_medhsapenet.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('48','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Jonske, Enrico Nasca Moon Kim Frederic<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=880\">Why does my medical AI look at pictures of birds? Exploring the efficacy of transfer learning across domain boundaries<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">Computer Methods and Programs in Biomedicine, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_49\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('49','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_49\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('49','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=86#tppubs\" title=\"Show all publications which have a relationship to this tag\">fine-tuning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=87#tppubs\" title=\"Show all publications which have a relationship to this tag\">foundation models<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=88#tppubs\" title=\"Show all publications which have a relationship to this tag\">Medical AI<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_49\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{jonske2025why,<br \/>\r\ntitle = {Why does my medical AI look at pictures of birds? Exploring the efficacy of transfer learning across domain boundaries},<br \/>\r\nauthor = {Enrico Nasca Moon Kim Frederic Jonske},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/jonske_whydoesmymedicalAIlookatpicturesofbirds.pdf},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-01-31},<br \/>\r\nurldate = {2025-01-31},<br \/>\r\njournal = {Computer Methods and Programs in Biomedicine},<br \/>\r\npublisher = {Elsevier},<br \/>\r\nkeywords = {deep learning, fine-tuning, foundation models, Medical AI},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('49','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_49\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/jonske_whydoesmymedicalAIlookatpicturesofbirds.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/jonske_whydoesmymedicalAIlook[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2025\/03\/jonske_whydoesmymedicalAIlook[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('49','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Han, Ting;  Adilova, Linara;  Petzka, Henning;  Kleesiek, Jens;  Kamp, Michael<\/p><p class=\"tp_pub_title\">Flatness is necessary, neural collapse is not: Rethinking generalization via grokking <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">Advances in Neural Information Processing Systems, <\/span><span class=\"tp_pub_additional_volume\">vol. 38, <\/span><span class=\"tp_pub_additional_pages\">pp. 106766\u2013106801, <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_50\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('50','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_50\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{han2026flatness,<br \/>\r\ntitle = {Flatness is necessary, neural collapse is not: Rethinking generalization via grokking},<br \/>\r\nauthor = {Ting Han and Linara Adilova and Henning Petzka and Jens Kleesiek and Michael Kamp},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-01-01},<br \/>\r\njournal = {Advances in Neural Information Processing Systems},<br \/>\r\nvolume = {38},<br \/>\r\npages = {106766\u2013106801},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('50','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Briq, Rania;  Kamp, Michael;  Fried, Ohad;  Cohen, Sarel;  Kesselheim, Stefan<\/p><p class=\"tp_pub_title\">The Amazing Stability of Flow Matching <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">EurIPS 2025 Workshop on Principles of Generative Modeling (PriGM), <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_53\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('53','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_53\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{briq2025amazing,<br \/>\r\ntitle = {The Amazing Stability of Flow Matching},<br \/>\r\nauthor = {Rania Briq and Michael Kamp and Ohad Fried and Sarel Cohen and Stefan Kesselheim},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-01-01},<br \/>\r\nbooktitle = {EurIPS 2025 Workshop on Principles of Generative Modeling (PriGM)},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('53','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Mian, Osman;  Kleesiek, Jens;  Kamp, Michael<\/p><p class=\"tp_pub_title\">Identifying Causal Sources from Missing Data <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">EurIPS Workshop on Causality for Impact (CausalityImpact), <\/span><span class=\"tp_pub_additional_year\">2025<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_54\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('54','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_54\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{main2025identifying,<br \/>\r\ntitle = {Identifying Causal Sources from Missing Data},<br \/>\r\nauthor = {Osman Mian and Jens Kleesiek and Michael Kamp},<br \/>\r\nyear  = {2025},<br \/>\r\ndate = {2025-01-01},<br \/>\r\nbooktitle = {EurIPS Workshop on Causality for Impact (CausalityImpact)},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('54','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2024\">2024<\/h3><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Salazer, Thomas L;  Sheth, Naitik;  Masud, Avais;  Serur, David;  Hidalgo, Guillermo;  Aqeel, Iram;  Adilova, Linara;  Kamp, Michael;  Fitzpatrick, Tim;  Krishnan, Sriram;  Rao, Kanishka;  Rao, Bharat<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=854\">Artificial Intelligence (AI)-Driven Screening for Undiscovered CKD<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">Journal of the American Society of Nephrology, <\/span><span class=\"tp_pub_additional_volume\">vol. 35, <\/span><span class=\"tp_pub_additional_issue\">iss. 10S, <\/span><span class=\"tp_pub_additional_pages\">pp. 10.1681, <\/span><span class=\"tp_pub_additional_year\">2024<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_44\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('44','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=76#tppubs\" title=\"Show all publications which have a relationship to this tag\">CKD<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=77#tppubs\" title=\"Show all publications which have a relationship to this tag\">healthcare<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=78#tppubs\" title=\"Show all publications which have a relationship to this tag\">medicine<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=79#tppubs\" title=\"Show all publications which have a relationship to this tag\">nephrology<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_44\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{salazer2024artificial,<br \/>\r\ntitle = {Artificial Intelligence (AI)-Driven Screening for Undiscovered CKD},<br \/>\r\nauthor = {Thomas L Salazer and Naitik Sheth and Avais Masud and David Serur and Guillermo Hidalgo and Iram Aqeel and Linara Adilova and Michael Kamp and Tim Fitzpatrick and Sriram Krishnan and Kanishka Rao and Bharat Rao},<br \/>\r\nyear  = {2024},<br \/>\r\ndate = {2024-10-01},<br \/>\r\nurldate = {2024-10-01},<br \/>\r\njournal = {Journal of the American Society of Nephrology},<br \/>\r\nvolume = {35},<br \/>\r\nissue = {10S},<br \/>\r\npages = {10.1681},<br \/>\r\npublisher = {LWW},<br \/>\r\nkeywords = {CKD, healthcare, medicine, nephrology},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('44','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Artificial Intelligence (AI)-Driven Screening for Undiscovered CKD\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/L2024Artificial-1.png\" width=\"160\" alt=\"Artificial Intelligence (AI)-Driven Screening for Undiscovered CKD\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Singh, Sidak Pal;  Adilova, Linara;  Kamp, Michael;  Fischer, Asja;  Sch\u00f6lkopf, Bernhard;  Hofmann, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=852\">Landscaping Linear Mode Connectivity<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">ICML Workshop on High-dimensional Learning Dynamics: The Emergence of Structure and Reasoning, <\/span><span class=\"tp_pub_additional_year\">2024<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_46\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('46','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=68#tppubs\" title=\"Show all publications which have a relationship to this tag\">linear mode connectivity<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=49#tppubs\" title=\"Show all publications which have a relationship to this tag\">theory of deep learning<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_46\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{singh2024landscaping,<br \/>\r\ntitle = {Landscaping Linear Mode Connectivity},<br \/>\r\nauthor = {Sidak Pal Singh and Linara Adilova and Michael Kamp and Asja Fischer and Bernhard Sch\u00f6lkopf and Thomas Hofmann},<br \/>\r\nyear  = {2024},<br \/>\r\ndate = {2024-09-01},<br \/>\r\nurldate = {2024-09-01},<br \/>\r\nbooktitle = {ICML Workshop on High-dimensional Learning Dynamics: The Emergence of Structure and Reasoning},<br \/>\r\nkeywords = {deep learning, linear mode connectivity, theory of deep learning},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('46','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Landscaping Linear Mode Connectivity\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Pal2024Landscaping-1.png\" width=\"160\" alt=\"Landscaping Linear Mode Connectivity\" \/><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Chen, Siming;  Gou, Liang;  Kamp, Michael;  Sunr, Dong<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=850\">Visual Computing for Autonomous Driving<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">IEEE Computer Graphics and Applications, <\/span><span class=\"tp_pub_additional_volume\">vol. 44, <\/span><span class=\"tp_pub_additional_issue\">iss. 3, <\/span><span class=\"tp_pub_additional_pages\">pp. 11-13, <\/span><span class=\"tp_pub_additional_year\">2024<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_45\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('45','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_45\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{chen2024visual,<br \/>\r\ntitle = {Visual Computing for Autonomous Driving},<br \/>\r\nauthor = {Siming Chen and Liang Gou and Michael Kamp and Dong Sunr},<br \/>\r\nyear  = {2024},<br \/>\r\ndate = {2024-06-21},<br \/>\r\nurldate = {2024-06-21},<br \/>\r\njournal = {IEEE Computer Graphics and Applications},<br \/>\r\nvolume = {44},<br \/>\r\nissue = {3},<br \/>\r\npages = {11-13},<br \/>\r\npublisher = {IEEE},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('45','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Visual Computing for Autonomous Driving\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Chen2024Visual-1.png\" width=\"160\" alt=\"Visual Computing for Autonomous Driving\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Adilova, Linara;  Andriushchenko, Maksym;  Fischer, Michael Kamp Asja;  Jaggi, Martin<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=847\">Layer-wise Linear Mode Connectivity<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">International Conference on Learning Representations (ICLR), <\/span><span class=\"tp_pub_additional_publisher\">Curran Associates, Inc, <\/span><span class=\"tp_pub_additional_year\">2024<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_40\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('40','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_40\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('40','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_40\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('40','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=67#tppubs\" title=\"Show all publications which have a relationship to this tag\">layer-wise<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=68#tppubs\" title=\"Show all publications which have a relationship to this tag\">linear mode connectivity<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_40\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{adilova2024layerwise,<br \/>\r\ntitle = {Layer-wise Linear Mode Connectivity},<br \/>\r\nauthor = {Linara Adilova and Maksym Andriushchenko and Michael Kamp Asja Fischer and Martin Jaggi},<br \/>\r\nurl = {https:\/\/openreview.net\/pdf?id=LfmZh91tDI},<br \/>\r\nyear  = {2024},<br \/>\r\ndate = {2024-05-07},<br \/>\r\nurldate = {2024-05-07},<br \/>\r\nbooktitle = {International Conference on Learning Representations (ICLR)},<br \/>\r\npublisher = {Curran Associates, Inc},<br \/>\r\nabstract = {Averaging neural network parameters is an intuitive method for fusing the knowledge of two independent models. It is most prominently used in federated learning. If models are averaged at the end of training, this can only lead to a good performing model if the loss surface of interest is very particular, i.e., the loss in the exact middle between the two models needs to be sufficiently low. This is impossible to guarantee for the non-convex losses of state-of-the-art networks. For averaging models trained on vastly different datasets, it was proposed to average only the parameters of particular layers or combinations of layers, resulting in better performing models. To get a better understanding of the effect of layer-wise averaging, we analyse the performance of the models that result from averaging single layers, or groups of layers. Based on our empirical and theoretical investigation, we introduce a novel notion of the layer-wise linear connectivity, and show that deep networks do not have layer-wise barriers between them. We analyze additionally the layer-wise personalization averaging and conjecture that in particular problem setup all the partial aggregations result in the approximately same performance.},<br \/>\r\nkeywords = {deep learning, layer-wise, linear mode connectivity},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('40','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_40\" style=\"display:none;\"><div class=\"tp_abstract_entry\">Averaging neural network parameters is an intuitive method for fusing the knowledge of two independent models. It is most prominently used in federated learning. If models are averaged at the end of training, this can only lead to a good performing model if the loss surface of interest is very particular, i.e., the loss in the exact middle between the two models needs to be sufficiently low. This is impossible to guarantee for the non-convex losses of state-of-the-art networks. For averaging models trained on vastly different datasets, it was proposed to average only the parameters of particular layers or combinations of layers, resulting in better performing models. To get a better understanding of the effect of layer-wise averaging, we analyse the performance of the models that result from averaging single layers, or groups of layers. Based on our empirical and theoretical investigation, we introduce a novel notion of the layer-wise linear connectivity, and show that deep networks do not have layer-wise barriers between them. We analyze additionally the layer-wise personalization averaging and conjecture that in particular problem setup all the partial aggregations result in the approximately same performance.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('40','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_40\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-globe\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/openreview.net\/pdf?id=LfmZh91tDI\" title=\"https:\/\/openreview.net\/pdf?id=LfmZh91tDI\" target=\"_blank\">https:\/\/openreview.net\/pdf?id=LfmZh91tDI<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('40','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Layer-wise Linear Mode Connectivity\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2024Layerwise.png\" width=\"160\" alt=\"Layer-wise Linear Mode Connectivity\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Yang, Fan;  Bodic, Pierre Le;  Kamp, Michael;  Boley, Mario<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=846\">Orthogonal Gradient Boosting for Interpretable Additive Rule Ensembles<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the 26th International Conference on Artificial Intelligence and Statistics (AISTATS), <\/span><span class=\"tp_pub_additional_year\">2024<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_41\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('41','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_41\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('41','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_41\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('41','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=69#tppubs\" title=\"Show all publications which have a relationship to this tag\">complexity<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=70#tppubs\" title=\"Show all publications which have a relationship to this tag\">explainability<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=71#tppubs\" title=\"Show all publications which have a relationship to this tag\">interpretability<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=63#tppubs\" title=\"Show all publications which have a relationship to this tag\">interpretable<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=72#tppubs\" title=\"Show all publications which have a relationship to this tag\">rule ensemble<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=73#tppubs\" title=\"Show all publications which have a relationship to this tag\">rule mining<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=65#tppubs\" title=\"Show all publications which have a relationship to this tag\">XAI<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_41\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{yang2024orthogonal,<br \/>\r\ntitle = {Orthogonal Gradient Boosting for Interpretable Additive Rule Ensembles},<br \/>\r\nauthor = {Fan Yang and Pierre Le Bodic and Michael Kamp and Mario Boley},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2024\/12\/yang24b.pdf},<br \/>\r\nyear  = {2024},<br \/>\r\ndate = {2024-05-02},<br \/>\r\nurldate = {2024-05-02},<br \/>\r\nbooktitle = {Proceedings of the 26th International Conference on Artificial Intelligence and Statistics (AISTATS)},<br \/>\r\nabstract = {Gradient boosting of prediction rules is an efficient approach to learn potentially interpretable yet accurate probabilistic models. However, actual interpretability requires to limit the number and size of the generated rules, and existing boosting variants are not designed for this purpose. Though corrective boosting refits all rule weights in each iteration to minimise prediction risk, the included rule conditions tend to be sub-optimal, because commonly used objective functions fail to anticipate this refitting. Here, we address this issue by a new objective function that measures the angle between the risk gradient vector and the projection of the condition output vector onto the orthogonal complement of the already selected conditions. This approach correctly approximates the ideal update of adding the risk gradient itself to the model and favours the inclusion of more general and thus shorter rules. As we demonstrate using a wide range of prediction tasks, this significantly improves the comprehensibility\/accuracy trade-off of the fitted ensemble. Additionally, we show how objective values for related rule conditions can be computed incrementally to avoid any substantial computational overhead of the new method.},<br \/>\r\nkeywords = {complexity, explainability, interpretability, interpretable, machine learning, rule ensemble, rule mining, XAI},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('41','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_41\" style=\"display:none;\"><div class=\"tp_abstract_entry\">Gradient boosting of prediction rules is an efficient approach to learn potentially interpretable yet accurate probabilistic models. However, actual interpretability requires to limit the number and size of the generated rules, and existing boosting variants are not designed for this purpose. Though corrective boosting refits all rule weights in each iteration to minimise prediction risk, the included rule conditions tend to be sub-optimal, because commonly used objective functions fail to anticipate this refitting. Here, we address this issue by a new objective function that measures the angle between the risk gradient vector and the projection of the condition output vector onto the orthogonal complement of the already selected conditions. This approach correctly approximates the ideal update of adding the risk gradient itself to the model and favours the inclusion of more general and thus shorter rules. As we demonstrate using a wide range of prediction tasks, this significantly improves the comprehensibility\/accuracy trade-off of the fitted ensemble. Additionally, we show how objective values for related rule conditions can be computed incrementally to avoid any substantial computational overhead of the new method.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('41','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_41\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2024\/12\/yang24b.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2024\/12\/yang24b.pdf\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2024\/12\/yang24b.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('41','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Orthogonal Gradient Boosting for Interpretable Additive Rule Ensembles\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Yang2024Orthogonal.png\" width=\"160\" alt=\"Orthogonal Gradient Boosting for Interpretable Additive Rule Ensembles\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2023\">2023<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Adilova, Linara;  Abourayya, Amr;  Li, Jianning;  Dada, Amin;  Petzka, Henning;  Egger, Jan;  Kleesiek, Jens;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=698\">FAM: Relative Flatness Aware Minimization<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the ICML Workshop on Topology, Algebra, and Geometry in Machine Learning (TAG-ML), <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_37\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('37','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_37\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('37','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=29#tppubs\" title=\"Show all publications which have a relationship to this tag\">flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=31#tppubs\" title=\"Show all publications which have a relationship to this tag\">generalization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=50#tppubs\" title=\"Show all publications which have a relationship to this tag\">relative flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=49#tppubs\" title=\"Show all publications which have a relationship to this tag\">theory of deep learning<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_37\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{adilova2023fam,<br \/>\r\ntitle = {FAM: Relative Flatness Aware Minimization},<br \/>\r\nauthor = {Linara Adilova and Amr Abourayya and Jianning Li and Amin Dada and Henning Petzka and Jan Egger and Jens Kleesiek and Michael Kamp},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/06\/fam_regularization.pdf},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-07-22},<br \/>\r\nurldate = {2023-07-22},<br \/>\r\nbooktitle = {Proceedings of the ICML Workshop on Topology, Algebra, and Geometry in Machine Learning (TAG-ML)},<br \/>\r\nkeywords = {deep learning, flatness, generalization, machine learning, relative flatness, theory of deep learning},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('37','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_37\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/06\/fam_regularization.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/06\/fam_regularization.pdf\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/06\/fam_regularization.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('37','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"FAM: Relative Flatness Aware Minimization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2023FAM.png\" width=\"160\" alt=\"FAM: Relative Flatness Aware Minimization\" \/><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\">Michael Kamp Linara Adilova, Gennady Andrienko<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=700\">Re-interpreting Rules Interpretability<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">International Journal of Data Science and Analytics, <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_38\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('38','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=63#tppubs\" title=\"Show all publications which have a relationship to this tag\">interpretable<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=66#tppubs\" title=\"Show all publications which have a relationship to this tag\">rule learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=65#tppubs\" title=\"Show all publications which have a relationship to this tag\">XAI<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_38\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{adilova2023reinterpreting,<br \/>\r\ntitle = {Re-interpreting Rules Interpretability},<br \/>\r\nauthor = {Linara Adilova, Michael Kamp, Gennady Andrienko, Natalia Andrienko},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-06-30},<br \/>\r\nurldate = {2023-06-30},<br \/>\r\njournal = {International Journal of Data Science and Analytics},<br \/>\r\nkeywords = {interpretable, machine learning, rule learning, XAI},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('38','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Re-interpreting Rules Interpretability\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2023Reinterpreting.png\" width=\"160\" alt=\"Re-interpreting Rules Interpretability\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Fischer, Jonas;  Vreeken, Jilles<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=701\">Federated Learning from Small Datasets<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">International Conference on Learning Representations (ICLR), <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_29\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('29','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_29\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('29','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=18#tppubs\" title=\"Show all publications which have a relationship to this tag\">black-box<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=41#tppubs\" title=\"Show all publications which have a relationship to this tag\">black-box parallelization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=55#tppubs\" title=\"Show all publications which have a relationship to this tag\">daisy<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=57#tppubs\" title=\"Show all publications which have a relationship to this tag\">daisy-chaining<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=58#tppubs\" title=\"Show all publications which have a relationship to this tag\">FedDC<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=53#tppubs\" title=\"Show all publications which have a relationship to this tag\">small<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=54#tppubs\" title=\"Show all publications which have a relationship to this tag\">small datasets<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_29\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2023federated,<br \/>\r\ntitle = {Federated Learning from Small Datasets},<br \/>\r\nauthor = {Michael Kamp and Jonas Fischer and Jilles Vreeken},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/08\/FederatedLearingSmallDatasets.pdf},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-05-01},<br \/>\r\nurldate = {2023-05-01},<br \/>\r\nbooktitle = {International Conference on Learning Representations (ICLR)},<br \/>\r\njournal = {arXiv preprint arXiv:2110.03469},<br \/>\r\nkeywords = {black-box, black-box parallelization, daisy, daisy-chaining, FedDC, federated learning, small, small datasets},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('29','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_29\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/08\/FederatedLearingSmallDatasets.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/08\/FederatedLearingSmallDatasets[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/08\/FederatedLearingSmallDatasets[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('29','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><a href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2023Federated.png\" target=\"_blank\"><img decoding=\"async\" name=\"Federated Learning from Small Datasets\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2023Federated.png\" width=\"160\" alt=\"Federated Learning from Small Datasets\" \/><\/a><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\">David Kaltenpoth Osman Mian, Michael Kamp<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=703\">Nothing but Regrets - Privacy-Preserving Federated Causal Discovery<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">International Conference on Artificial Intelligence and Statistics (AISTATS), <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_31\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('31','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=59#tppubs\" title=\"Show all publications which have a relationship to this tag\">causal discovery<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=60#tppubs\" title=\"Show all publications which have a relationship to this tag\">causality<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=62#tppubs\" title=\"Show all publications which have a relationship to this tag\">explainable<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=19#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=61#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated causal discovery<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=63#tppubs\" title=\"Show all publications which have a relationship to this tag\">interpretable<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_31\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{mian2022nothing,<br \/>\r\ntitle = {Nothing but Regrets - Privacy-Preserving Federated Causal Discovery},<br \/>\r\nauthor = {Osman Mian, David Kaltenpoth, Michael Kamp, Jilles Vreeken},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-04-25},<br \/>\r\nurldate = {2023-04-25},<br \/>\r\nbooktitle = {International Conference on Artificial Intelligence and Statistics (AISTATS)},<br \/>\r\nkeywords = {causal discovery, causality, explainable, federated, federated causal discovery, federated learning, interpretable},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('31','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Nothing but Regrets - Privacy-Preserving Federated Causal Discovery\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Mian2023Nothing.png\" width=\"160\" alt=\"Nothing but Regrets - Privacy-Preserving Federated Causal Discovery\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Mian, Osman;  Kamp, Michael;  Vreeken, Jilles<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=706\">Information-Theoretic Causal Discovery and Intervention Detection over Multiple Environments<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_33\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('33','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=59#tppubs\" title=\"Show all publications which have a relationship to this tag\">causal discovery<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=60#tppubs\" title=\"Show all publications which have a relationship to this tag\">causality<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=19#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=61#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated causal discovery<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=64#tppubs\" title=\"Show all publications which have a relationship to this tag\">intervention<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_33\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{mian2023informationb,<br \/>\r\ntitle = {Information-Theoretic Causal Discovery and Intervention Detection over Multiple Environments},<br \/>\r\nauthor = {Osman Mian and Michael Kamp and Jilles Vreeken},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-02-07},<br \/>\r\nurldate = {2023-02-07},<br \/>\r\nbooktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},<br \/>\r\nkeywords = {causal discovery, causality, federated, federated causal discovery, federated learning, intervention},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('33','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Information-Theoretic Causal Discovery and Intervention Detection over Multiple Environments\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Mian2023InformationTheoretic.png\" width=\"160\" alt=\"Information-Theoretic Causal Discovery and Intervention Detection over Multiple Environments\" \/><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Li, Jianning;  Ferreira, Andr\u00e9;  Puladi, Behrus;  Alves, Victor;  Kamp, Michael;  Kim, Moon;  Nensa, Felix;  Kleesiek, Jens;  Ahmadi, Seyed-Ahmad;  Egger, Jan<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=708\">Open-source skull reconstruction with MONAI<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">SoftwareX, <\/span><span class=\"tp_pub_additional_volume\">vol. 23, <\/span><span class=\"tp_pub_additional_pages\">pp. 101432, <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_36\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('36','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_36\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{li2023open,<br \/>\r\ntitle = {Open-source skull reconstruction with MONAI},<br \/>\r\nauthor = {Jianning Li and Andr\u00e9 Ferreira and Behrus Puladi and Victor Alves and Michael Kamp and Moon Kim and Felix Nensa and Jens Kleesiek and Seyed-Ahmad Ahmadi and Jan Egger},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-01-01},<br \/>\r\nurldate = {2023-01-01},<br \/>\r\njournal = {SoftwareX},<br \/>\r\nvolume = {23},<br \/>\r\npages = {101432},<br \/>\r\npublisher = {Elsevier},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('36','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Open-source skull reconstruction with MONAI\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Li2023Opensource.png\" width=\"160\" alt=\"Open-source skull reconstruction with MONAI\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Adilova, Linara;  Chen, Siming;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=707\">Informed Novelty Detection in Sequential Data by Per-Cluster Modeling<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">ICML workshop on Artificial Intelligence &amp; Human Computer Interaction, <\/span><span class=\"tp_pub_additional_year\">2023<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_39\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('39','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_39\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('39','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_39\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{adilova2023informed,<br \/>\r\ntitle = {Informed Novelty Detection in Sequential Data by Per-Cluster Modeling},<br \/>\r\nauthor = {Linara Adilova and Siming Chen and Michael Kamp},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/09\/Informed_Novelty_Detection_in_Sequential_Data_by_Per_Cluster_Modeling.pdf},<br \/>\r\nyear  = {2023},<br \/>\r\ndate = {2023-01-01},<br \/>\r\nurldate = {2023-01-01},<br \/>\r\nbooktitle = {ICML workshop on Artificial Intelligence & Human Computer Interaction},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('39','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_39\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/09\/Informed_Novelty_Detection_in_Sequential_Data_by_Per_Cluster_Modeling.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/09\/Informed_Novelty_Detection_in[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2023\/09\/Informed_Novelty_Detection_in[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('39','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Informed Novelty Detection in Sequential Data by Per-Cluster Modeling\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2023Informed.png\" width=\"160\" alt=\"Informed Novelty Detection in Sequential Data by Per-Cluster Modeling\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2022\">2022<\/h3><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Wang, Junhong;  Li, Yun;  Zhou, Zhaoyu;  Wang, Chengshun;  Hou, Yijie;  Zhang, Li;  Xue, Xiangyang;  Kamp, Michael;  Zhang, Xiaolong;  Chen, Siming<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=713\">When, Where and How does it fail? A Spatial-temporal Visual Analytics Approach for Interpretable Object Detection in Autonomous Driving<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">IEEE Transactions on Visualization and Computer Graphics, <\/span><span class=\"tp_pub_additional_year\">2022<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_34\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('34','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_34\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{wang2022and,<br \/>\r\ntitle = {When, Where and How does it fail? A Spatial-temporal Visual Analytics Approach for Interpretable Object Detection in Autonomous Driving},<br \/>\r\nauthor = {Junhong Wang and Yun Li and Zhaoyu Zhou and Chengshun Wang and Yijie Hou and Li Zhang and Xiangyang Xue and Michael Kamp and Xiaolong Zhang and Siming Chen},<br \/>\r\nyear  = {2022},<br \/>\r\ndate = {2022-01-01},<br \/>\r\nurldate = {2022-01-01},<br \/>\r\njournal = {IEEE Transactions on Visualization and Computer Graphics},<br \/>\r\npublisher = {IEEE},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('34','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"When, Where and How does it fail? A Spatial-temporal Visual Analytics Approach for Interpretable Object Detection in Autonomous Driving\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Wang2022When.png\" width=\"160\" alt=\"When, Where and How does it fail? A Spatial-temporal Visual Analytics Approach for Interpretable Object Detection in Autonomous Driving\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Mian, Osman;  Kaltenpoth, David;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=709\">Regret-based Federated Causal Discovery<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">The KDD&#039;22 Workshop on Causal Discovery, <\/span><span class=\"tp_pub_additional_pages\">pp. 61\u201369, <\/span><span class=\"tp_pub_additional_organization\">PMLR <\/span><span class=\"tp_pub_additional_year\">2022<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_35\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('35','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_35\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{mian2022regret,<br \/>\r\ntitle = {Regret-based Federated Causal Discovery},<br \/>\r\nauthor = {Osman Mian and David Kaltenpoth and Michael Kamp},<br \/>\r\nyear  = {2022},<br \/>\r\ndate = {2022-01-01},<br \/>\r\nurldate = {2022-01-01},<br \/>\r\nbooktitle = {The KDD&#039;22 Workshop on Causal Discovery},<br \/>\r\npages = {61--69},<br \/>\r\norganization = {PMLR},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('35','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Regret-based Federated Causal Discovery\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Mian2022Regretbased.png\" width=\"160\" alt=\"Regret-based Federated Causal Discovery\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2021\">2021<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Petzka, Henning;  Kamp, Michael;  Adilova, Linara;  Sminchisescu, Cristian;  Boley, Mario<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=714\">Relative Flatness and Generalization<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Advances in Neural Information Processing Systems, <\/span><span class=\"tp_pub_additional_publisher\">Curran Associates, Inc., <\/span><span class=\"tp_pub_additional_year\">2021<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_26\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('26','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_26\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('26','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=29#tppubs\" title=\"Show all publications which have a relationship to this tag\">flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=31#tppubs\" title=\"Show all publications which have a relationship to this tag\">generalization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=51#tppubs\" title=\"Show all publications which have a relationship to this tag\">Hessian<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=20#tppubs\" title=\"Show all publications which have a relationship to this tag\">learning theory<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=50#tppubs\" title=\"Show all publications which have a relationship to this tag\">relative flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=49#tppubs\" title=\"Show all publications which have a relationship to this tag\">theory of deep learning<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_26\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{petzka2021relative,<br \/>\r\ntitle = {Relative Flatness and Generalization},<br \/>\r\nauthor = {Henning Petzka and Michael Kamp and Linara Adilova and Cristian Sminchisescu and Mario Boley},<br \/>\r\nyear  = {2021},<br \/>\r\ndate = {2021-12-07},<br \/>\r\nurldate = {2021-12-07},<br \/>\r\nbooktitle = {Advances in Neural Information Processing Systems},<br \/>\r\npublisher = {Curran Associates, Inc.},<br \/>\r\nabstract = {Flatness of the loss curve is conjectured to be connected to the generalization ability of machine learning models, in particular neural networks. While it has been empirically observed that flatness measures consistently correlate strongly with generalization, it is still an open theoretical problem why and under which circumstances flatness is connected to generalization, in particular in light of reparameterizations that change certain flatness measures but leave generalization unchanged. We investigate the connection between flatness and generalization by relating it to the interpolation from representative data, deriving notions of representativeness, and feature robustness. The notions allow us to rigorously connect flatness and generalization and to identify conditions under which the connection holds. Moreover, they give rise to a novel, but natural relative flatness measure that correlates strongly with generalization, simplifies to ridge regression for ordinary least squares, and solves the reparameterization issue.},<br \/>\r\nkeywords = {deep learning, flatness, generalization, Hessian, learning theory, relative flatness, theory of deep learning},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('26','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_26\" style=\"display:none;\"><div class=\"tp_abstract_entry\">Flatness of the loss curve is conjectured to be connected to the generalization ability of machine learning models, in particular neural networks. While it has been empirically observed that flatness measures consistently correlate strongly with generalization, it is still an open theoretical problem why and under which circumstances flatness is connected to generalization, in particular in light of reparameterizations that change certain flatness measures but leave generalization unchanged. We investigate the connection between flatness and generalization by relating it to the interpolation from representative data, deriving notions of representativeness, and feature robustness. The notions allow us to rigorously connect flatness and generalization and to identify conditions under which the connection holds. Moreover, they give rise to a novel, but natural relative flatness measure that correlates strongly with generalization, simplifies to ridge regression for ordinary least squares, and solves the reparameterization issue.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('26','tp_abstract')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Relative Flatness and Generalization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Petzka2021Relative.png\" width=\"160\" alt=\"Relative Flatness and Generalization\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Linsner, Florian;  Adilova, Linara;  D\u00e4ubener, Sina;  Kamp, Michael;  Fischer, Asja<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=715\">Approaches to Uncertainty Quantification in Federated Deep Learning<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Machine Learning and Principles and Practice of Knowledge Discovery in Databases: International Workshops of ECML PKDD 2021, <\/span><span class=\"tp_pub_additional_volume\">vol. 2, <\/span><span class=\"tp_pub_additional_publisher\">Springer, <\/span><span class=\"tp_pub_additional_year\">2021<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_27\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('27','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_27\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('27','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=52#tppubs\" title=\"Show all publications which have a relationship to this tag\">uncertainty<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_27\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{linsner2021uncertainty,<br \/>\r\ntitle = {Approaches to Uncertainty Quantification in Federated Deep Learning},<br \/>\r\nauthor = {Florian Linsner and Linara Adilova and Sina D\u00e4ubener and Michael Kamp and Asja Fischer},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/04\/federatedUncertainty.pdf},<br \/>\r\nyear  = {2021},<br \/>\r\ndate = {2021-09-17},<br \/>\r\nurldate = {2021-09-17},<br \/>\r\nbooktitle = {Machine Learning and Principles and Practice of Knowledge Discovery in Databases: International Workshops of ECML PKDD 2021},<br \/>\r\nissuetitle = {Workshop on Parallel, Distributed, and Federated Learning},<br \/>\r\nvolume = {2},<br \/>\r\npages = {128-145},<br \/>\r\npublisher = {Springer},<br \/>\r\nkeywords = {federated learning, uncertainty},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('27','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_27\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/04\/federatedUncertainty.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/04\/federatedUncertainty.pdf\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2022\/04\/federatedUncertainty.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('27','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Approaches to Uncertainty Quantification in Federated Deep Learning\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Linsner2021Approaches.png\" width=\"160\" alt=\"Approaches to Uncertainty Quantification in Federated Deep Learning\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Li, Xiaoxiao;  Jiang, Meirui;  Zhang, Xiaofei;  Kamp, Michael;  Dou, Qi<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=716\">FedBN: Federated Learning on Non-IID Features via Local Batch Normalization<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the 9th International Conference on Learning Representations (ICLR), <\/span><span class=\"tp_pub_additional_year\">2021<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_25\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('25','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_25\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('25','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_25\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('25','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=48#tppubs\" title=\"Show all publications which have a relationship to this tag\">batch normalization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=41#tppubs\" title=\"Show all publications which have a relationship to this tag\">black-box parallelization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_25\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{li2021fedbn,<br \/>\r\ntitle = {FedBN: Federated Learning on Non-IID Features via Local Batch Normalization},<br \/>\r\nauthor = {Xiaoxiao Li and Meirui Jiang and Xiaofei Zhang and Michael Kamp and Qi Dou},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/fedbn_federated_learning_on_non_iid_features_via_local_batch_normalization.pdf<br \/>\r\nhttps:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/FedBN_appendix.pdf},<br \/>\r\nyear  = {2021},<br \/>\r\ndate = {2021-05-03},<br \/>\r\nurldate = {2021-05-03},<br \/>\r\nbooktitle = {Proceedings of the 9th International Conference on Learning Representations (ICLR)},<br \/>\r\nabstract = {The emerging paradigm of federated learning (FL) strives to enable collaborative training of deep models on the network edge without centrally aggregating raw data and hence improving data privacy. In most cases, the assumption of independent and identically distributed samples across local clients does not hold for federated learning setups. Under this setting, neural network training performance may vary significantly according to the data distribution and even hurt training convergence. Most of the previous work has focused on a difference in the distribution of labels or client shifts. Unlike those settings, we address an important problem of FL, e.g., different scanners\/sensors in medical imaging, different scenery distribution in autonomous driving (highway vs. city), where local clients store examples with different distributions compared to other clients, which we denote as feature shift non-iid. In this work, we propose an effective method that uses local batch normalization to alleviate the feature shift before averaging models. The resulting scheme, called FedBN, outperforms both classical FedAvg, as well as the state-of-the-art for non-iid data (FedProx) on our extensive experiments. These empirical results are supported by a convergence analysis that shows in a simplified setting that FedBN has a faster convergence rate than FedAvg. Code is available at https:\/\/github.com\/med-air\/FedBN.},<br \/>\r\nkeywords = {batch normalization, black-box parallelization, deep learning, federated learning},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('25','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_25\" style=\"display:none;\"><div class=\"tp_abstract_entry\">The emerging paradigm of federated learning (FL) strives to enable collaborative training of deep models on the network edge without centrally aggregating raw data and hence improving data privacy. In most cases, the assumption of independent and identically distributed samples across local clients does not hold for federated learning setups. Under this setting, neural network training performance may vary significantly according to the data distribution and even hurt training convergence. Most of the previous work has focused on a difference in the distribution of labels or client shifts. Unlike those settings, we address an important problem of FL, e.g., different scanners\/sensors in medical imaging, different scenery distribution in autonomous driving (highway vs. city), where local clients store examples with different distributions compared to other clients, which we denote as feature shift non-iid. In this work, we propose an effective method that uses local batch normalization to alleviate the feature shift before averaging models. The resulting scheme, called FedBN, outperforms both classical FedAvg, as well as the state-of-the-art for non-iid data (FedProx) on our extensive experiments. These empirical results are supported by a convergence analysis that shows in a simplified setting that FedBN has a faster convergence rate than FedAvg. Code is available at https:\/\/github.com\/med-air\/FedBN.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('25','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_25\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/fedbn_federated_learning_on_non_iid_features_via_local_batch_normalization.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/fedbn_federated_learning_on_n[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/fedbn_federated_learning_on_n[...]<\/a><\/li><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/FedBN_appendix.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/FedBN_appendix.pdf\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2021\/05\/FedBN_appendix.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('25','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"FedBN: Federated Learning on Non-IID Features via Local Batch Normalization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Li2021FedBN.png\" width=\"160\" alt=\"FedBN: Federated Learning on Non-IID Features via Local Batch Normalization\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2020\">2020<\/h3><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Heppe, Lukas;  Kamp, Michael;  Adilova, Linara;  Piatkowski, Nico;  Heinrich, Danny;  Morik, Katharina<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=717\">Resource-Constrained On-Device Learning by Dynamic Averaging<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Proceedings of the Workshop on Parallel, Distributed, and Federated Learning (PDFL) at ECMLPKDD, <\/span><span class=\"tp_pub_additional_year\">2020<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_24\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('24','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_24\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('24','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_24\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('24','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=41#tppubs\" title=\"Show all publications which have a relationship to this tag\">black-box parallelization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=43#tppubs\" title=\"Show all publications which have a relationship to this tag\">distributed learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=47#tppubs\" title=\"Show all publications which have a relationship to this tag\">edge computing<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=45#tppubs\" title=\"Show all publications which have a relationship to this tag\">embedded<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=42#tppubs\" title=\"Show all publications which have a relationship to this tag\">exponential family<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=46#tppubs\" title=\"Show all publications which have a relationship to this tag\">FPGA<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=44#tppubs\" title=\"Show all publications which have a relationship to this tag\">resource-efficient<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_24\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{heppe2020resource,<br \/>\r\ntitle = {Resource-Constrained On-Device Learning by Dynamic Averaging},<br \/>\r\nauthor = {Lukas Heppe and Michael Kamp and Linara Adilova and Nico Piatkowski and Danny Heinrich and Katharina Morik},<br \/>\r\nurl = {https:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/10\/Resource_Constrained_Federated_Learning-1.pdf},<br \/>\r\nyear  = {2020},<br \/>\r\ndate = {2020-09-14},<br \/>\r\nurldate = {2020-09-14},<br \/>\r\nbooktitle = {Proceedings of the Workshop on Parallel, Distributed, and Federated Learning (PDFL) at ECMLPKDD},<br \/>\r\nabstract = {The communication between data-generating devices is partially responsible for a growing portion of the world\u2019s power consumption. Thus reducing communication is vital, both, from an economical and an ecological perspective. For machine learning, on-device learning avoids sending raw data, which can reduce communication substantially. Furthermore, not centralizing the data protects privacy-sensitive data. However, most learning algorithms require hardware with high computation power and thus high energy consumption. In contrast, ultra-low-power processors, like FPGAs or micro-controllers, allow for energy-efficient learning of local models. Combined with communication-efficient distributed learning strategies, this reduces the overall energy consumption and enables applications that were yet impossible due to limited energy on local devices. The major challenge is then, that the low-power processors typically only have integer processing capabilities. This paper investigates an approach to communication-efficient on-device learning of integer exponential families that can be executed on low-power processors, is privacy-preserving, and effectively minimizes communication. The empirical evaluation shows that the approach can reach a model quality comparable to a centrally learned regular model with an order of magnitude less communication. Comparing the overall energy consumption, this reduces the required energy for solving the machine learning task by a significant amount.},<br \/>\r\nkeywords = {black-box parallelization, distributed learning, edge computing, embedded, exponential family, FPGA, resource-efficient},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('24','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_24\" style=\"display:none;\"><div class=\"tp_abstract_entry\">The communication between data-generating devices is partially responsible for a growing portion of the world\u2019s power consumption. Thus reducing communication is vital, both, from an economical and an ecological perspective. For machine learning, on-device learning avoids sending raw data, which can reduce communication substantially. Furthermore, not centralizing the data protects privacy-sensitive data. However, most learning algorithms require hardware with high computation power and thus high energy consumption. In contrast, ultra-low-power processors, like FPGAs or micro-controllers, allow for energy-efficient learning of local models. Combined with communication-efficient distributed learning strategies, this reduces the overall energy consumption and enables applications that were yet impossible due to limited energy on local devices. The major challenge is then, that the low-power processors typically only have integer processing capabilities. This paper investigates an approach to communication-efficient on-device learning of integer exponential families that can be executed on low-power processors, is privacy-preserving, and effectively minimizes communication. The empirical evaluation shows that the approach can reach a model quality comparable to a centrally learned regular model with an order of magnitude less communication. Comparing the overall energy consumption, this reduces the required energy for solving the machine learning task by a significant amount.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('24','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_24\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/10\/Resource_Constrained_Federated_Learning-1.pdf\" title=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/10\/Resource_Constrained_Federate[...]\" target=\"_blank\">https:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/10\/Resource_Constrained_Federate[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('24','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Resource-Constrained On-Device Learning by Dynamic Averaging\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Heppe2020ResourceConstrained.png\" width=\"160\" alt=\"Resource-Constrained On-Device Learning by Dynamic Averaging\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Petzka, Henning;  Adilova, Linara;  Kamp, Michael;  Sminchisescu, Cristian<\/p><p class=\"tp_pub_title\"><a href=\"\">Feature-Robustness, Flatness and Generalization Error for Deep Neural Networks<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_year\">2020<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_16\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('16','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_16\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('16','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=29#tppubs\" title=\"Show all publications which have a relationship to this tag\">flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=31#tppubs\" title=\"Show all publications which have a relationship to this tag\">generalization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=20#tppubs\" title=\"Show all publications which have a relationship to this tag\">learning theory<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=28#tppubs\" title=\"Show all publications which have a relationship to this tag\">loss surface<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=27#tppubs\" title=\"Show all publications which have a relationship to this tag\">neural networks<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=30#tppubs\" title=\"Show all publications which have a relationship to this tag\">robustness<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_16\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{petzka2020feature,<br \/>\r\ntitle = {Feature-Robustness, Flatness and Generalization Error for Deep Neural Networks},<br \/>\r\nauthor = {Henning Petzka and Linara Adilova and Michael Kamp and Cristian Sminchisescu},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/01\/flatnessFeatureRobustnessGeneralization.pdf},<br \/>\r\nyear  = {2020},<br \/>\r\ndate = {2020-01-01},<br \/>\r\nurldate = {2020-01-01},<br \/>\r\njournal = {arXiv preprint arXiv:2001.00939},<br \/>\r\nkeywords = {deep learning, flatness, generalization, learning theory, loss surface, neural networks, robustness},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('16','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_16\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/01\/flatnessFeatureRobustnessGeneralization.pdf\" title=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/01\/flatnessFeatureRobustnessGener[...]\" target=\"_blank\">http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/01\/flatnessFeatureRobustnessGener[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('16','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Welke, Pascal;  Seiffarth, Florian;  Kamp, Michael;  Wrobel, Stefan<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=718\">HOPS: Probabilistic Subtree Mining for Small and Large Graphs<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery &amp;amp; Data Mining, <\/span><span class=\"tp_pub_additional_pages\">pp. 1275\u20131284, <\/span><span class=\"tp_pub_additional_publisher\">Association for Computing Machinery, <\/span><span class=\"tp_pub_additional_address\">Virtual Event, CA, USA, <\/span><span class=\"tp_pub_additional_year\">2020<\/span>, <span class=\"tp_pub_additional_isbn\">ISBN: 9781450379984<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_23\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('23','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_23\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('23','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_23\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('23','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_23\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{10.1145\/3394486.3403180,<br \/>\r\ntitle = {HOPS: Probabilistic Subtree Mining for Small and Large Graphs},<br \/>\r\nauthor = {Pascal Welke and Florian Seiffarth and Michael Kamp and Stefan Wrobel},<br \/>\r\nurl = {https:\/\/doi.org\/10.1145\/3394486.3403180},<br \/>\r\ndoi = {10.1145\/3394486.3403180},<br \/>\r\nisbn = {9781450379984},<br \/>\r\nyear  = {2020},<br \/>\r\ndate = {2020-01-01},<br \/>\r\nurldate = {2020-01-01},<br \/>\r\nbooktitle = {Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery &amp; Data Mining},<br \/>\r\npages = {1275\u20131284},<br \/>\r\npublisher = {Association for Computing Machinery},<br \/>\r\naddress = {Virtual Event, CA, USA},<br \/>\r\nseries = {KDD &#039;20},<br \/>\r\nabstract = {Frequent subgraph mining, i.e., the identification of relevant patterns in graph databases, is a well-known data mining problem with high practical relevance, since next to summarizing the data, the resulting patterns can also be used to define powerful domain-specific similarity functions for prediction. In recent years, significant progress has been made towards subgraph mining algorithms that scale to complex graphs by focusing on tree patterns and probabilistically allowing a small amount of incompleteness in the result. Nonetheless, the complexity of the pattern matching component used for deciding subtree isomorphism on arbitrary graphs has significantly limited the scalability of existing approaches. In this paper, we adapt sampling techniques from mathematical combinatorics to the problem of probabilistic subtree mining in arbitrary databases of many small to medium-size graphs or a single large graph. By restricting on tree patterns, we provide an algorithm that approximately counts or decides subtree isomorphism for arbitrary transaction graphs in sub-linear time with one-sided error. Our empirical evaluation on a range of benchmark graph datasets shows that the novel algorithm substantially outperforms state-of-the-art approaches both in the task of approximate counting of embeddings in single large graphs and in probabilistic frequent subtree mining in large databases of small to medium sized graphs.},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('23','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_23\" style=\"display:none;\"><div class=\"tp_abstract_entry\">Frequent subgraph mining, i.e., the identification of relevant patterns in graph databases, is a well-known data mining problem with high practical relevance, since next to summarizing the data, the resulting patterns can also be used to define powerful domain-specific similarity functions for prediction. In recent years, significant progress has been made towards subgraph mining algorithms that scale to complex graphs by focusing on tree patterns and probabilistically allowing a small amount of incompleteness in the result. Nonetheless, the complexity of the pattern matching component used for deciding subtree isomorphism on arbitrary graphs has significantly limited the scalability of existing approaches. In this paper, we adapt sampling techniques from mathematical combinatorics to the problem of probabilistic subtree mining in arbitrary databases of many small to medium-size graphs or a single large graph. By restricting on tree patterns, we provide an algorithm that approximately counts or decides subtree isomorphism for arbitrary transaction graphs in sub-linear time with one-sided error. Our empirical evaluation on a range of benchmark graph datasets shows that the novel algorithm substantially outperforms state-of-the-art approaches both in the task of approximate counting of embeddings in single large graphs and in probabilistic frequent subtree mining in large databases of small to medium sized graphs.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('23','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_23\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-globe\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/doi.org\/10.1145\/3394486.3403180\" title=\"https:\/\/doi.org\/10.1145\/3394486.3403180\" target=\"_blank\">https:\/\/doi.org\/10.1145\/3394486.3403180<\/a><\/li><li><i class=\"ai ai-doi\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/dx.doi.org\/10.1145\/3394486.3403180\" title=\"Follow DOI:10.1145\/3394486.3403180\" target=\"_blank\">doi:10.1145\/3394486.3403180<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('23','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"HOPS: Probabilistic Subtree Mining for Small and Large Graphs\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Welke2020HOPS.png\" width=\"160\" alt=\"HOPS: Probabilistic Subtree Mining for Small and Large Graphs\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2019\">2019<\/h3><div class=\"tp_publication tp_publication_phdthesis\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=719\">Black-Box Parallelization for Machine Learning<\/a> <span class=\"tp_pub_type tp_  phdthesis\">PhD Thesis<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_school\">Universit\u00e4ts-und Landesbibliothek Bonn, <\/span><span class=\"tp_pub_additional_year\">2019<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_17\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('17','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_17\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('17','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_17\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('17','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=22#tppubs\" title=\"Show all publications which have a relationship to this tag\">averaging<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=18#tppubs\" title=\"Show all publications which have a relationship to this tag\">black-box<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=25#tppubs\" title=\"Show all publications which have a relationship to this tag\">communication-efficient<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=21#tppubs\" title=\"Show all publications which have a relationship to this tag\">convex optimization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=7#tppubs\" title=\"Show all publications which have a relationship to this tag\">distributed<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=24#tppubs\" title=\"Show all publications which have a relationship to this tag\">dynamic averaging<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=19#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=20#tppubs\" title=\"Show all publications which have a relationship to this tag\">learning theory<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=4#tppubs\" title=\"Show all publications which have a relationship to this tag\">parallelization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=26#tppubs\" title=\"Show all publications which have a relationship to this tag\">privacy<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=23#tppubs\" title=\"Show all publications which have a relationship to this tag\">radon machine<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_17\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@phdthesis{kamp2019black,<br \/>\r\ntitle = {Black-Box Parallelization for Machine Learning},<br \/>\r\nauthor = {Michael Kamp},<br \/>\r\nurl = {https:\/\/d-nb.info\/1200020057\/34},<br \/>\r\nyear  = {2019},<br \/>\r\ndate = {2019-01-01},<br \/>\r\nurldate = {2019-01-01},<br \/>\r\nschool = {Universit\u00e4ts-und Landesbibliothek Bonn},<br \/>\r\nabstract = {The landscape of machine learning applications is changing rapidly: large centralized datasets are replaced by high volume, high velocity data streams generated by a vast number of geographically distributed, loosely connected devices, such as mobile phones, smart sensors, autonomous vehicles or industrial machines. Current learning approaches centralize the data and process it in parallel in a cluster or computing center. This has three major disadvantages: (i) it does not scale well with the number of data-generating devices since their growth exceeds that of computing centers, (ii) the communication costs for centralizing the data are prohibitive in many applications, and (iii) it requires sharing potentially privacy-sensitive data. Pushing computation towards the data-generating devices alleviates these problems and allows to employ their otherwise unused computing power. However, current parallel learning approaches are designed for tightly integrated systems with low latency and high bandwidth, not for loosely connected distributed devices. Therefore, I propose a new paradigm for parallelization that treats the learning algorithm as a black box, training local models on distributed devices and aggregating them into a single strong one. Since this requires only exchanging models instead of actual data, the approach is highly scalable, communication-efficient, and privacy-preserving.<br \/>\r\nFollowing this paradigm, this thesis develops black-box parallelizations for two broad classes of learning algorithms. One approach can be applied to incremental learning algorithms, i.e., those that improve a model in iterations. Based on the utility of aggregations it schedules communication dynamically, adapting it to the hardness of the learning problem. In practice, this leads to a reduction in communication by orders of magnitude. It is analyzed for (i) online learning, in particular in the context of in-stream learning, which allows to guarantee optimal regret and for (ii) batch learning based on empirical risk minimization where optimal convergence can be guaranteed. The other approach is applicable to non-incremental algorithms as well. It uses a novel aggregation method based on the Radon point that allows to achieve provably high model quality with only a single aggregation. This is achieved in polylogarithmic runtime on quasi-polynomially many processors. This relates parallel machine learning to Nick\u2019s class of parallel decision problems and is a step towards answering a fundamental open problem about the abilities and limitations of efficient parallel learning algorithms. An empirical study on real distributed systems confirms the potential of the approaches in realistic application scenarios.},<br \/>\r\nkeywords = {averaging, black-box, communication-efficient, convex optimization, deep learning, distributed, dynamic averaging, federated, learning theory, machine learning, parallelization, privacy, radon machine},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {phdthesis}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('17','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_17\" style=\"display:none;\"><div class=\"tp_abstract_entry\">The landscape of machine learning applications is changing rapidly: large centralized datasets are replaced by high volume, high velocity data streams generated by a vast number of geographically distributed, loosely connected devices, such as mobile phones, smart sensors, autonomous vehicles or industrial machines. Current learning approaches centralize the data and process it in parallel in a cluster or computing center. This has three major disadvantages: (i) it does not scale well with the number of data-generating devices since their growth exceeds that of computing centers, (ii) the communication costs for centralizing the data are prohibitive in many applications, and (iii) it requires sharing potentially privacy-sensitive data. Pushing computation towards the data-generating devices alleviates these problems and allows to employ their otherwise unused computing power. However, current parallel learning approaches are designed for tightly integrated systems with low latency and high bandwidth, not for loosely connected distributed devices. Therefore, I propose a new paradigm for parallelization that treats the learning algorithm as a black box, training local models on distributed devices and aggregating them into a single strong one. Since this requires only exchanging models instead of actual data, the approach is highly scalable, communication-efficient, and privacy-preserving.<br \/>\r\nFollowing this paradigm, this thesis develops black-box parallelizations for two broad classes of learning algorithms. One approach can be applied to incremental learning algorithms, i.e., those that improve a model in iterations. Based on the utility of aggregations it schedules communication dynamically, adapting it to the hardness of the learning problem. In practice, this leads to a reduction in communication by orders of magnitude. It is analyzed for (i) online learning, in particular in the context of in-stream learning, which allows to guarantee optimal regret and for (ii) batch learning based on empirical risk minimization where optimal convergence can be guaranteed. The other approach is applicable to non-incremental algorithms as well. It uses a novel aggregation method based on the Radon point that allows to achieve provably high model quality with only a single aggregation. This is achieved in polylogarithmic runtime on quasi-polynomially many processors. This relates parallel machine learning to Nick\u2019s class of parallel decision problems and is a step towards answering a fundamental open problem about the abilities and limitations of efficient parallel learning algorithms. An empirical study on real distributed systems confirms the potential of the approaches in realistic application scenarios.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('17','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_17\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-globe\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/d-nb.info\/1200020057\/34\" title=\"https:\/\/d-nb.info\/1200020057\/34\" target=\"_blank\">https:\/\/d-nb.info\/1200020057\/34<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('17','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Black-Box Parallelization for Machine Learning\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2019BlackBox.png\" width=\"160\" alt=\"Black-Box Parallelization for Machine Learning\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Adilova, Linara;  Natious, Livin;  Chen, Siming;  Thonnard, Olivier;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=721\">System Misuse Detection via Informed Behavior Clustering and Modeling<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">2019 49th Annual IEEE\/IFIP International Conference on Dependable Systems and Networks Workshops (DSN-W), <\/span><span class=\"tp_pub_additional_organization\">IEEE <\/span><span class=\"tp_pub_additional_year\">2019<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_18\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('18','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_18\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('18','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=35#tppubs\" title=\"Show all publications which have a relationship to this tag\">anomaly detection<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=33#tppubs\" title=\"Show all publications which have a relationship to this tag\">cybersecurity<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=13#tppubs\" title=\"Show all publications which have a relationship to this tag\">DiSIEM<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=34#tppubs\" title=\"Show all publications which have a relationship to this tag\">security<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=36#tppubs\" title=\"Show all publications which have a relationship to this tag\">user behavior modelling<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=11#tppubs\" title=\"Show all publications which have a relationship to this tag\">visualization<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_18\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{adilova2019system,<br \/>\r\ntitle = {System Misuse Detection via Informed Behavior Clustering and Modeling},<br \/>\r\nauthor = {Linara Adilova and Livin Natious and Siming Chen and Olivier Thonnard and Michael Kamp},<br \/>\r\nurl = {https:\/\/arxiv.org\/pdf\/1907.00874},<br \/>\r\nyear  = {2019},<br \/>\r\ndate = {2019-01-01},<br \/>\r\nurldate = {2019-01-01},<br \/>\r\nbooktitle = {2019 49th Annual IEEE\/IFIP International Conference on Dependable Systems and Networks Workshops (DSN-W)},<br \/>\r\npages = {15--23},<br \/>\r\norganization = {IEEE},<br \/>\r\nkeywords = {anomaly detection, cybersecurity, DiSIEM, security, user behavior modelling, visualization},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('18','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_18\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"ai ai-arxiv\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/arxiv.org\/pdf\/1907.00874\" title=\"https:\/\/arxiv.org\/pdf\/1907.00874\" target=\"_blank\">https:\/\/arxiv.org\/pdf\/1907.00874<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('18','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"System Misuse Detection via Informed Behavior Clustering and Modeling\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2019System.png\" width=\"160\" alt=\"System Misuse Detection via Informed Behavior Clustering and Modeling\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Petzka, Henning;  Adilova, Linara;  Kamp, Michael;  Sminchisescu, Cristian<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=702\">A Reparameterization-Invariant Flatness Measure for Deep Neural Networks<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Science meets Engineering of Deep Learning workshop at NeurIPS, <\/span><span class=\"tp_pub_additional_year\">2019<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_19\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('19','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_19\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('19','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=29#tppubs\" title=\"Show all publications which have a relationship to this tag\">flatness<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=31#tppubs\" title=\"Show all publications which have a relationship to this tag\">generalization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=20#tppubs\" title=\"Show all publications which have a relationship to this tag\">learning theory<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=28#tppubs\" title=\"Show all publications which have a relationship to this tag\">loss surface<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=27#tppubs\" title=\"Show all publications which have a relationship to this tag\">neural networks<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=30#tppubs\" title=\"Show all publications which have a relationship to this tag\">robustness<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_19\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{petzka2019reparameterization,<br \/>\r\ntitle = {A Reparameterization-Invariant Flatness Measure for Deep Neural Networks},<br \/>\r\nauthor = {Henning Petzka and Linara Adilova and Michael Kamp and Cristian Sminchisescu},<br \/>\r\nurl = {https:\/\/arxiv.org\/pdf\/1912.00058},<br \/>\r\nyear  = {2019},<br \/>\r\ndate = {2019-01-01},<br \/>\r\nurldate = {2019-01-01},<br \/>\r\nbooktitle = {Science meets Engineering of Deep Learning workshop at NeurIPS},<br \/>\r\nkeywords = {deep learning, flatness, generalization, learning theory, loss surface, neural networks, robustness},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('19','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_19\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"ai ai-arxiv\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/arxiv.org\/pdf\/1912.00058\" title=\"https:\/\/arxiv.org\/pdf\/1912.00058\" target=\"_blank\">https:\/\/arxiv.org\/pdf\/1912.00058<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('19','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"A Reparameterization-Invariant Flatness Measure for Deep Neural Networks\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Petzka2019A.png\" width=\"160\" alt=\"A Reparameterization-Invariant Flatness Measure for Deep Neural Networks\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Adilova, Linara;  Rosenzweig, Julia;  Kamp, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=720\">Information Theoretic Perspective of Federated Learning<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">NeurIPS Workshop on Information Theory and Machine Learning, <\/span><span class=\"tp_pub_additional_year\">2019<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_20\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('20','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_20\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('20','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_20\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{adilova2019information,<br \/>\r\ntitle = {Information Theoretic Perspective of Federated Learning},<br \/>\r\nauthor = {Linara Adilova and Julia Rosenzweig and Michael Kamp},<br \/>\r\nurl = {https:\/\/arxiv.org\/pdf\/1911.07652},<br \/>\r\nyear  = {2019},<br \/>\r\ndate = {2019-01-01},<br \/>\r\nurldate = {2019-01-01},<br \/>\r\nbooktitle = {NeurIPS Workshop on Information Theory and Machine Learning},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('20','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_20\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"ai ai-arxiv\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/arxiv.org\/pdf\/1911.07652\" title=\"https:\/\/arxiv.org\/pdf\/1911.07652\" target=\"_blank\">https:\/\/arxiv.org\/pdf\/1911.07652<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('20','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Information Theoretic Perspective of Federated Learning\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Adilova2019InformationTheoretic.png\" width=\"160\" alt=\"Information Theoretic Perspective of Federated Learning\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2018\">2018<\/h3><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Giesselbach, Sven;  Ullrich, Katrin;  Kamp, Michael;  Paurat, Daniel;  G\u00e4rtner, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=722\">Corresponding Projections for Orphan Screening<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Proceedings of the ML4H workshop at NeurIPS, <\/span><span class=\"tp_pub_additional_year\">2018<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_14\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('14','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_14\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('14','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=15#tppubs\" title=\"Show all publications which have a relationship to this tag\">corresponding projections<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=2#tppubs\" title=\"Show all publications which have a relationship to this tag\">transfer learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=3#tppubs\" title=\"Show all publications which have a relationship to this tag\">unsupervised<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_14\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{giesselbach2018corresponding,<br \/>\r\ntitle = {Corresponding Projections for Orphan Screening},<br \/>\r\nauthor = {Sven Giesselbach and Katrin Ullrich and Michael Kamp and Daniel Paurat and Thomas G\u00e4rtner},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/12\/cpNIPS.pdf},<br \/>\r\nyear  = {2018},<br \/>\r\ndate = {2018-12-08},<br \/>\r\nurldate = {2018-12-08},<br \/>\r\nbooktitle = {Proceedings of the ML4H workshop at NeurIPS},<br \/>\r\nkeywords = {corresponding projections, transfer learning, unsupervised},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('14','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_14\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/12\/cpNIPS.pdf\" title=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/12\/cpNIPS.pdf\" target=\"_blank\">http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/12\/cpNIPS.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('14','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Corresponding Projections for Orphan Screening\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Giesselbach2018Corresponding.png\" width=\"160\" alt=\"Corresponding Projections for Orphan Screening\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Nguyen, Phong H.;  Chen, Siming;  Andrienko, Natalia;  Kamp, Michael;  Adilova, Linara;  Andrienko, Gennady;  Thonnard, Olivier;  Bessani, Alysson;  Turkay, Cagatay<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=723\">Designing Visualisation Enhancements for SIEM Systems<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">15th IEEE Symposium on Visualization for Cyber Security \u2013 VizSec, <\/span><span class=\"tp_pub_additional_year\">2018<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_13\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('13','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_13\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('13','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=13#tppubs\" title=\"Show all publications which have a relationship to this tag\">DiSIEM<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=12#tppubs\" title=\"Show all publications which have a relationship to this tag\">SIEM<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=14#tppubs\" title=\"Show all publications which have a relationship to this tag\">visual analytics<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=11#tppubs\" title=\"Show all publications which have a relationship to this tag\">visualization<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_13\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{phong2018designing,<br \/>\r\ntitle = {Designing Visualisation Enhancements for SIEM Systems},<br \/>\r\nauthor = {Phong H. Nguyen and Siming Chen and Natalia Andrienko and Michael Kamp and Linara Adilova and Gennady Andrienko and Olivier Thonnard and Alysson Bessani and Cagatay Turkay},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/vizsec2018-poster-designing-visualisation-enhancements-for-siem-systems\/},<br \/>\r\nyear  = {2018},<br \/>\r\ndate = {2018-10-22},<br \/>\r\nurldate = {2018-10-22},<br \/>\r\nbooktitle = {15th IEEE Symposium on Visualization for Cyber Security \u2013 VizSec},<br \/>\r\nkeywords = {DiSIEM, SIEM, visual analytics, visualization},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('13','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_13\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-globe\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/vizsec2018-poster-designing-visualisation-enhancements-for-siem-systems\/\" title=\"http:\/\/michaelkamp.org\/vizsec2018-poster-designing-visualisation-enhancements-fo[...]\" target=\"_blank\">http:\/\/michaelkamp.org\/vizsec2018-poster-designing-visualisation-enhancements-fo[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('13','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Designing Visualisation Enhancements for SIEM Systems\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/H2018Designing.png\" width=\"160\" alt=\"Designing Visualisation Enhancements for SIEM Systems\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Adilova, Linara;  Sicking, Joachim;  H\u00fcger, Fabian;  Schlicht, Peter;  Wirtz, Tim;  Wrobel, Stefan<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=724\">Efficient Decentralized Deep Learning by Dynamic Model Averaging<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Machine Learning and Knowledge Discovery in Databases, <\/span><span class=\"tp_pub_additional_publisher\">Springer, <\/span><span class=\"tp_pub_additional_year\">2018<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_12\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('12','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_12\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('12','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_12\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('12','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=5#tppubs\" title=\"Show all publications which have a relationship to this tag\">decentralized<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=10#tppubs\" title=\"Show all publications which have a relationship to this tag\">deep learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_12\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2018efficient,<br \/>\r\ntitle = {Efficient Decentralized Deep Learning by Dynamic Model Averaging},<br \/>\r\nauthor = {Michael Kamp and Linara Adilova and Joachim Sicking and Fabian H\u00fcger and Peter Schlicht and Tim Wirtz and Stefan Wrobel},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/07\/commEffDeepLearning_extended.pdf},<br \/>\r\nyear  = {2018},<br \/>\r\ndate = {2018-09-14},<br \/>\r\nurldate = {2018-09-14},<br \/>\r\nbooktitle = {Machine Learning and Knowledge Discovery in Databases},<br \/>\r\npublisher = {Springer},<br \/>\r\nabstract = {We propose an efficient protocol for decentralized training of deep neural networks from distributed data sources. The proposed protocol allows to handle different phases of model training equally well and to quickly adapt to concept drifts. This leads to a reduction of communication by an order of magnitude compared to periodically communicating state-of-the-art approaches. Moreover, we derive a communication bound that scales well with the hardness of the serialized learning problem. The reduction in communication comes at almost no cost, as the predictive performance remains virtually unchanged. Indeed, the proposed protocol retains loss bounds of periodically averaging schemes. An extensive empirical evaluation validates major improvement of the trade-off between model performance and communication which could be beneficial for numerous decentralized learning applications, such as autonomous driving, or voice recognition and image classification on mobile phones.},<br \/>\r\nkeywords = {decentralized, deep learning, federated learning},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('12','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_12\" style=\"display:none;\"><div class=\"tp_abstract_entry\">We propose an efficient protocol for decentralized training of deep neural networks from distributed data sources. The proposed protocol allows to handle different phases of model training equally well and to quickly adapt to concept drifts. This leads to a reduction of communication by an order of magnitude compared to periodically communicating state-of-the-art approaches. Moreover, we derive a communication bound that scales well with the hardness of the serialized learning problem. The reduction in communication comes at almost no cost, as the predictive performance remains virtually unchanged. Indeed, the proposed protocol retains loss bounds of periodically averaging schemes. An extensive empirical evaluation validates major improvement of the trade-off between model performance and communication which could be beneficial for numerous decentralized learning applications, such as autonomous driving, or voice recognition and image classification on mobile phones.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('12','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_12\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/07\/commEffDeepLearning_extended.pdf\" title=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/07\/commEffDeepLearning_extended.p[...]\" target=\"_blank\">http:\/\/michaelkamp.org\/wp-content\/uploads\/2018\/07\/commEffDeepLearning_extended.p[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('12','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Efficient Decentralized Deep Learning by Dynamic Model Averaging\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2018Efficient.png\" width=\"160\" alt=\"Efficient Decentralized Deep Learning by Dynamic Model Averaging\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2017\">2017<\/h3><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Gunar Ernis, Michael Kamp<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=129\">Machine Learning f\u00fcr die smarte Produktion<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">VDMA-Nachrichten, <\/span><span class=\"tp_pub_additional_pages\">pp. 36-37, <\/span><span class=\"tp_pub_additional_year\">2017<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_15\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('15','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_15\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('15','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=17#tppubs\" title=\"Show all publications which have a relationship to this tag\">industry 4.0<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=16#tppubs\" title=\"Show all publications which have a relationship to this tag\">smart production<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_15\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{kamp2017machine,<br \/>\r\ntitle = {Machine Learning f\u00fcr die smarte Produktion},<br \/>\r\nauthor = {Gunar Ernis, Michael Kamp},<br \/>\r\neditor = {Rebecca Pini},<br \/>\r\nurl = {https:\/\/sud.vdma.org\/documents\/15012668\/22571546\/VDMA-Nachrichten%20Smart%20Data%2011-2017_1513086481204.pdf\/c5767569-504e-4f64-8dba-8e7bdd06c18e},<br \/>\r\nyear  = {2017},<br \/>\r\ndate = {2017-11-01},<br \/>\r\nissuetitle = {Smart Data - aus Daten Gold machen},<br \/>\r\njournal = {VDMA-Nachrichten},<br \/>\r\npages = {36-37},<br \/>\r\npublisher = {Verband Deutscher Maschinen- und Anlagenbau e.V.},<br \/>\r\nkeywords = {industry 4.0, machine learning, smart production},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('15','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_15\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-globe\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/sud.vdma.org\/documents\/15012668\/22571546\/VDMA-Nachrichten%20Smart%20Data%2011-2017_1513086481204.pdf\/c5767569-504e-4f64-8dba-8e7bdd06c18e\" title=\"https:\/\/sud.vdma.org\/documents\/15012668\/22571546\/VDMA-Nachrichten%20Smart%20Data[...]\" target=\"_blank\">https:\/\/sud.vdma.org\/documents\/15012668\/22571546\/VDMA-Nachrichten%20Smart%20Data[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('15','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><\/div><\/div><div class=\"tp_publication tp_publication_article\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Flouris, Ioannis;  Giatrakos, Nikos;  Deligiannakis, Antonios;  Garofalakis, Minos;  Kamp, Michael;  Mock, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=737\">Issues in Complex Event Processing: Status and Prospects in the Big Data Era<\/a> <span class=\"tp_pub_type tp_  article\">Journal Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_journal\">Journal of Systems and Software, <\/span><span class=\"tp_pub_additional_year\">2017<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_1\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('1','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_1\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@article{flouris2016issues,<br \/>\r\ntitle = {Issues in Complex Event Processing: Status and Prospects in the Big Data Era},<br \/>\r\nauthor = {Ioannis Flouris and Nikos Giatrakos and Antonios Deligiannakis and Minos Garofalakis and Michael Kamp and Michael Mock},<br \/>\r\nyear  = {2017},<br \/>\r\ndate = {2017-01-01},<br \/>\r\nurldate = {2017-01-01},<br \/>\r\njournal = {Journal of Systems and Software},<br \/>\r\npublisher = {Elsevier},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {article}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('1','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Issues in Complex Event Processing: Status and Prospects in the Big Data Era\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Flouris2017Issues.png\" width=\"160\" alt=\"Issues in Complex Event Processing: Status and Prospects in the Big Data Era\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  Missura, Olana;  G\u00e4rtner, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=726\">Effective Parallelisation for Machine Learning<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Advances in Neural Information Processing Systems, <\/span><span class=\"tp_pub_additional_pages\">pp. 6480\u20136491, <\/span><span class=\"tp_pub_additional_year\">2017<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_10\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('10','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_10\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('10','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_10\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('10','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=5#tppubs\" title=\"Show all publications which have a relationship to this tag\">decentralized<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=7#tppubs\" title=\"Show all publications which have a relationship to this tag\">distributed<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=8#tppubs\" title=\"Show all publications which have a relationship to this tag\">machine learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=4#tppubs\" title=\"Show all publications which have a relationship to this tag\">parallelization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=6#tppubs\" title=\"Show all publications which have a relationship to this tag\">radon<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_10\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2017effective,<br \/>\r\ntitle = {Effective Parallelisation for Machine Learning},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Olana Missura and Thomas G\u00e4rtner},<br \/>\r\nurl = {http:\/\/papers.nips.cc\/paper\/7226-effective-parallelisation-for-machine-learning.pdf},<br \/>\r\nyear  = {2017},<br \/>\r\ndate = {2017-01-01},<br \/>\r\nurldate = {2017-01-01},<br \/>\r\nbooktitle = {Advances in Neural Information Processing Systems},<br \/>\r\npages = {6480--6491},<br \/>\r\nabstract = {We present a novel parallelisation scheme that simplifies the adaptation of learning algorithms to growing amounts of data as well as growing needs for accurate and confident predictions in critical applications. In contrast to other parallelisation techniques, it can be applied to a broad class of learning algorithms without further mathematical derivations and without writing dedicated code, while at the same time maintaining theoretical performance guarantees. Moreover, our parallelisation scheme is able to reduce the runtime of many learning algorithms to polylogarithmic time on quasi-polynomially many processing units. This is a significant step towards a general answer to an open question on efficient parallelisation of machine learning algorithms in the sense of Nick&#039;s Class (NC). The cost of this parallelisation is in the form of a larger sample complexity. Our empirical study confirms the potential of our parallelisation scheme with fixed numbers of processors and instances in realistic application scenarios.},<br \/>\r\nkeywords = {decentralized, distributed, machine learning, parallelization, radon},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('10','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_10\" style=\"display:none;\"><div class=\"tp_abstract_entry\">We present a novel parallelisation scheme that simplifies the adaptation of learning algorithms to growing amounts of data as well as growing needs for accurate and confident predictions in critical applications. In contrast to other parallelisation techniques, it can be applied to a broad class of learning algorithms without further mathematical derivations and without writing dedicated code, while at the same time maintaining theoretical performance guarantees. Moreover, our parallelisation scheme is able to reduce the runtime of many learning algorithms to polylogarithmic time on quasi-polynomially many processing units. This is a significant step towards a general answer to an open question on efficient parallelisation of machine learning algorithms in the sense of Nick&#039;s Class (NC). The cost of this parallelisation is in the form of a larger sample complexity. Our empirical study confirms the potential of our parallelisation scheme with fixed numbers of processors and instances in realistic application scenarios.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('10','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_10\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/papers.nips.cc\/paper\/7226-effective-parallelisation-for-machine-learning.pdf\" title=\"http:\/\/papers.nips.cc\/paper\/7226-effective-parallelisation-for-machine-learning.[...]\" target=\"_blank\">http:\/\/papers.nips.cc\/paper\/7226-effective-parallelisation-for-machine-learning.[...]<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('10','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Effective Parallelisation for Machine Learning\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2017Effective.png\" width=\"160\" alt=\"Effective Parallelisation for Machine Learning\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Ullrich, Katrin;  Kamp, Michael;  G\u00e4rtner, Thomas;  Vogt, Martin;  Wrobel, Stefan<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=725\">Co-regularised support vector regression<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Joint European Conference on Machine Learning and Knowledge Discovery in Databases, <\/span><span class=\"tp_pub_additional_pages\">pp. 338\u2013354, <\/span><span class=\"tp_pub_additional_organization\">Springer <\/span><span class=\"tp_pub_additional_year\">2017<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_11\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('11','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_11\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('11','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=1#tppubs\" title=\"Show all publications which have a relationship to this tag\">co-regularization<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=2#tppubs\" title=\"Show all publications which have a relationship to this tag\">transfer learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=3#tppubs\" title=\"Show all publications which have a relationship to this tag\">unsupervised<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_11\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{ullrich2017co,<br \/>\r\ntitle = {Co-regularised support vector regression},<br \/>\r\nauthor = {Katrin Ullrich and Michael Kamp and Thomas G\u00e4rtner and Martin Vogt and Stefan Wrobel},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/mk_v1\/wp-content\/uploads\/2018\/05\/CoRegSVR.pdf},<br \/>\r\nyear  = {2017},<br \/>\r\ndate = {2017-01-01},<br \/>\r\nurldate = {2017-01-01},<br \/>\r\nbooktitle = {Joint European Conference on Machine Learning and Knowledge Discovery in Databases},<br \/>\r\npages = {338--354},<br \/>\r\norganization = {Springer},<br \/>\r\nkeywords = {co-regularization, transfer learning, unsupervised},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('11','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_11\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/mk_v1\/wp-content\/uploads\/2018\/05\/CoRegSVR.pdf\" title=\"http:\/\/michaelkamp.org\/mk_v1\/wp-content\/uploads\/2018\/05\/CoRegSVR.pdf\" target=\"_blank\">http:\/\/michaelkamp.org\/mk_v1\/wp-content\/uploads\/2018\/05\/CoRegSVR.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('11','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Co-regularised support vector regression\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Ullrich2017Coregularised.png\" width=\"160\" alt=\"Co-regularised support vector regression\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2016\">2016<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Bothe, Sebastian;  Boley, Mario;  Mock, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=727\">Communication-Efficient Distributed Online Learning with Kernels<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span> Frasconi, Paolo;  Landwehr, Niels;  Manco, Giuseppe;  Vreeken, Jilles (Ed.): <span class=\"tp_pub_additional_booktitle\">Machine Learning and Knowledge Discovery in Databases, <\/span><span class=\"tp_pub_additional_pages\">pp. 805\u2013819, <\/span><span class=\"tp_pub_additional_publisher\">Springer International Publishing, <\/span><span class=\"tp_pub_additional_year\">2016<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_21\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('21','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_21\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('21','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_21\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('21','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=25#tppubs\" title=\"Show all publications which have a relationship to this tag\">communication-efficient<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=7#tppubs\" title=\"Show all publications which have a relationship to this tag\">distributed<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=24#tppubs\" title=\"Show all publications which have a relationship to this tag\">dynamic averaging<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=9#tppubs\" title=\"Show all publications which have a relationship to this tag\">federated learning<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=37#tppubs\" title=\"Show all publications which have a relationship to this tag\">kernel methods<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=4#tppubs\" title=\"Show all publications which have a relationship to this tag\">parallelization<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_21\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2016communication,<br \/>\r\ntitle = {Communication-Efficient Distributed Online Learning with Kernels},<br \/>\r\nauthor = {Michael Kamp and Sebastian Bothe and Mario Boley and Michael Mock},<br \/>\r\neditor = {Paolo Frasconi and Niels Landwehr and Giuseppe Manco and Jilles Vreeken},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/Paper467.pdf},<br \/>\r\nyear  = {2016},<br \/>\r\ndate = {2016-09-16},<br \/>\r\nurldate = {2016-09-16},<br \/>\r\nbooktitle = {Machine Learning and Knowledge Discovery in Databases},<br \/>\r\npages = {805--819},<br \/>\r\npublisher = {Springer International Publishing},<br \/>\r\nabstract = {We propose an efficient distributed online learning protocol for low-latency real-time services. It extends a previously presented protocol to kernelized online learners that represent their models by a support vector expansion. While such learners often achieve higher predictive performance than their linear counterparts, communicating the support vector expansions becomes inefficient for large numbers of support vectors. The proposed extension allows for a larger class of online learning algorithms\u2014including those alleviating the problem above through model compression. In addition, we characterize the quality of the proposed protocol by introducing a novel criterion that requires the communication to be bounded by the loss suffered.},<br \/>\r\nkeywords = {communication-efficient, distributed, dynamic averaging, federated learning, kernel methods, parallelization},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('21','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_21\" style=\"display:none;\"><div class=\"tp_abstract_entry\">We propose an efficient distributed online learning protocol for low-latency real-time services. It extends a previously presented protocol to kernelized online learners that represent their models by a support vector expansion. While such learners often achieve higher predictive performance than their linear counterparts, communicating the support vector expansions becomes inefficient for large numbers of support vectors. The proposed extension allows for a larger class of online learning algorithms\u2014including those alleviating the problem above through model compression. In addition, we characterize the quality of the proposed protocol by introducing a novel criterion that requires the communication to be bounded by the loss suffered.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('21','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_21\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/Paper467.pdf\" title=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/Paper467.pdf\" target=\"_blank\">http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/Paper467.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('21','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Communication-Efficient Distributed Online Learning with Kernels\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2016Communicationefficient.png\" width=\"160\" alt=\"Communication-Efficient Distributed Online Learning with Kernels\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Ullrich, Katrin;  Kamp, Michael;  G\u00e4rtner, Thomas;  Vogt, Martin;  Wrobel, Stefan<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=728\">Ligand-based virtual screening with co-regularised support Vector Regression<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">2016 IEEE 16th international conference on data mining workshops (ICDMW), <\/span><span class=\"tp_pub_additional_pages\">pp. 261\u2013268, <\/span><span class=\"tp_pub_additional_organization\">IEEE <\/span><span class=\"tp_pub_additional_year\">2016<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_abstract_link\"><a id=\"tp_abstract_sh_22\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('22','tp_abstract')\" title=\"Show abstract\" style=\"cursor:pointer;\">Abstract<\/a><\/span> | <span class=\"tp_resource_link\"><a id=\"tp_links_sh_22\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('22','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_22\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('22','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=39#tppubs\" title=\"Show all publications which have a relationship to this tag\">biology<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=40#tppubs\" title=\"Show all publications which have a relationship to this tag\">chemistry<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=15#tppubs\" title=\"Show all publications which have a relationship to this tag\">corresponding projections<\/a>, <a rel=\"nofollow\" href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;tgid=38#tppubs\" title=\"Show all publications which have a relationship to this tag\">semi-supervised<\/a><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_22\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{ullrich2016ligand,<br \/>\r\ntitle = {Ligand-based virtual screening with co-regularised support Vector Regression},<br \/>\r\nauthor = {Katrin Ullrich and Michael Kamp and Thomas G\u00e4rtner and Martin Vogt and Stefan Wrobel},<br \/>\r\nurl = {http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/LigandBasedCoSVR.pdf},<br \/>\r\nyear  = {2016},<br \/>\r\ndate = {2016-01-01},<br \/>\r\nurldate = {2016-01-01},<br \/>\r\nbooktitle = {2016 IEEE 16th international conference on data mining workshops (ICDMW)},<br \/>\r\npages = {261--268},<br \/>\r\norganization = {IEEE},<br \/>\r\nabstract = {We consider the problem of ligand affinity prediction as a regression task, typically with few labelled examples, many unlabelled instances, and multiple views on the data. In chemoinformatics, the prediction of binding affinities for protein ligands is an important but also challenging task. As protein-ligand bonds trigger biochemical reactions, their characterisation is a crucial step in the process of drug discovery and design. However, the practical determination of ligand affinities is very expensive, whereas unlabelled compounds are available in abundance. Additionally, many different vectorial representations for compounds (molecular fingerprints) exist that cover different sets of features. To this task we propose to apply a co-regularisation approach, which extracts information from unlabelled examples by ensuring that individual models trained on different fingerprints make similar predictions. We extend support vector regression similarly to the existing co-regularised least squares regression (CoRLSR) and obtain a co-regularised support vector regression (CoSVR). We empirically evaluate the performance of CoSVR on various protein-ligand datasets. We show that CoSVR outperforms CoRLSR as well as existing state-of-the-art approaches that do not take unlabelled molecules into account. Additionally, we provide a theoretical bound on the Rademacher complexity for CoSVR.},<br \/>\r\nkeywords = {biology, chemistry, corresponding projections, semi-supervised},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('22','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_abstract\" id=\"tp_abstract_22\" style=\"display:none;\"><div class=\"tp_abstract_entry\">We consider the problem of ligand affinity prediction as a regression task, typically with few labelled examples, many unlabelled instances, and multiple views on the data. In chemoinformatics, the prediction of binding affinities for protein ligands is an important but also challenging task. As protein-ligand bonds trigger biochemical reactions, their characterisation is a crucial step in the process of drug discovery and design. However, the practical determination of ligand affinities is very expensive, whereas unlabelled compounds are available in abundance. Additionally, many different vectorial representations for compounds (molecular fingerprints) exist that cover different sets of features. To this task we propose to apply a co-regularisation approach, which extracts information from unlabelled examples by ensuring that individual models trained on different fingerprints make similar predictions. We extend support vector regression similarly to the existing co-regularised least squares regression (CoRLSR) and obtain a co-regularised support vector regression (CoSVR). We empirically evaluate the performance of CoSVR on various protein-ligand datasets. We show that CoSVR outperforms CoRLSR as well as existing state-of-the-art approaches that do not take unlabelled molecules into account. Additionally, we provide a theoretical bound on the Rademacher complexity for CoSVR.<\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('22','tp_abstract')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_22\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/LigandBasedCoSVR.pdf\" title=\"http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/LigandBasedCoSVR.pdf\" target=\"_blank\">http:\/\/michaelkamp.org\/wp-content\/uploads\/2020\/03\/LigandBasedCoSVR.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('22','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Ligand-based virtual screening with co-regularised support Vector Regression\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Ullrich2016Ligandbased.png\" width=\"160\" alt=\"Ligand-based virtual screening with co-regularised support Vector Regression\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2015\">2015<\/h3><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  G\u00e4rtner, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=729\">Parallelizing Randomized Convex Optimization<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Proceedings of the 8th NIPS Workshop on Optimization for Machine Learning, <\/span><span class=\"tp_pub_additional_year\">2015<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_2\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('2','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_2\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('2','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_2\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{kamp2015parallelizing,<br \/>\r\ntitle = {Parallelizing Randomized Convex Optimization},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Thomas G\u00e4rtner},<br \/>\r\nurl = {http:\/\/www.opt-ml.org\/papers\/OPT2015_paper_23.pdf},<br \/>\r\nyear  = {2015},<br \/>\r\ndate = {2015-01-01},<br \/>\r\nurldate = {2015-01-01},<br \/>\r\nbooktitle = {Proceedings of the 8th NIPS Workshop on Optimization for Machine Learning},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('2','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_2\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/www.opt-ml.org\/papers\/OPT2015_paper_23.pdf\" title=\"http:\/\/www.opt-ml.org\/papers\/OPT2015_paper_23.pdf\" target=\"_blank\">http:\/\/www.opt-ml.org\/papers\/OPT2015_paper_23.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('2','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Parallelizing Randomized Convex Optimization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2015Parallelizing.png\" width=\"160\" alt=\"Parallelizing Randomized Convex Optimization\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2014\">2014<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  G\u00e4rtner, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=731\">Beating Human Analysts in Nowcasting Corporate Earnings by using Publicly Available Stock Price and Correlation Features<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Proceedings of the SIAM International Conference on Data Mining, <\/span><span class=\"tp_pub_additional_pages\">pp. 641\u2013649, <\/span><span class=\"tp_pub_additional_organization\">SIAM <\/span><span class=\"tp_pub_additional_year\">2014<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_3\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('3','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_3\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('3','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_3\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{michael2014beating,<br \/>\r\ntitle = {Beating Human Analysts in Nowcasting Corporate Earnings by using Publicly Available Stock Price and Correlation Features},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Thomas G\u00e4rtner},<br \/>\r\nurl = {http:\/\/www.ferari-project.eu\/wp-content\/uploads\/2014\/12\/earningsPrediction.pdf},<br \/>\r\nyear  = {2014},<br \/>\r\ndate = {2014-01-01},<br \/>\r\nurldate = {2014-01-01},<br \/>\r\nbooktitle = {Proceedings of the SIAM International Conference on Data Mining},<br \/>\r\nvolume = {72},<br \/>\r\npages = {641--649},<br \/>\r\norganization = {SIAM},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('3','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_3\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"fas fa-file-pdf\"><\/i><a class=\"tp_pub_list\" href=\"http:\/\/www.ferari-project.eu\/wp-content\/uploads\/2014\/12\/earningsPrediction.pdf\" title=\"http:\/\/www.ferari-project.eu\/wp-content\/uploads\/2014\/12\/earningsPrediction.pdf\" target=\"_blank\">http:\/\/www.ferari-project.eu\/wp-content\/uploads\/2014\/12\/earningsPrediction.pdf<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('3','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Beating Human Analysts in Nowcasting Corporate Earnings by using Publicly Available Stock Price and Correlation Features\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2014Beating.png\" width=\"160\" alt=\"Beating Human Analysts in Nowcasting Corporate Earnings by using Publicly Available Stock Price and Correlation Features\" \/><\/div><\/div><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  Keren, Daniel;  Schuster, Assaf;  Sharfman, Izchak<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=732\">Communication-Efficient Distributed Online Prediction by Dynamic Model Synchronization<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">European Conference on Machine Learning and Principles and Practice of Knowledge Discovery (ECMLPKDD), <\/span><span class=\"tp_pub_additional_organization\">Springer <\/span><span class=\"tp_pub_additional_year\">2014<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_4\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('4','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_4\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2014communication,<br \/>\r\ntitle = {Communication-Efficient Distributed Online Prediction by Dynamic Model Synchronization},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Daniel Keren and Assaf Schuster and Izchak Sharfman},<br \/>\r\nyear  = {2014},<br \/>\r\ndate = {2014-01-01},<br \/>\r\nurldate = {2014-01-01},<br \/>\r\nbooktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery (ECMLPKDD)},<br \/>\r\norganization = {Springer},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('4','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Communication-Efficient Distributed Online Prediction by Dynamic Model Synchronization\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2014CommunicationEfficient.png\" width=\"160\" alt=\"Communication-Efficient Distributed Online Prediction by Dynamic Model Synchronization\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  Mock, Michael;  Keren, Daniel;  Schuster, Assaf;  Sharfman, Izchak<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=730\">Adaptive Communication Bounds for Distributed Online Learning<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">Proceedings of the 7th NIPS Workshop on Optimization for Machine Learning, <\/span><span class=\"tp_pub_additional_year\">2014<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_resource_link\"><a id=\"tp_links_sh_5\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('5','tp_links')\" title=\"Show links and resources\" style=\"cursor:pointer;\">Links<\/a><\/span> | <span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_5\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('5','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_5\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{kamp2014adaptive,<br \/>\r\ntitle = {Adaptive Communication Bounds for Distributed Online Learning},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Michael Mock and Daniel Keren and Assaf Schuster and Izchak Sharfman},<br \/>\r\nurl = {https:\/\/arxiv.org\/abs\/1911.12896},<br \/>\r\nyear  = {2014},<br \/>\r\ndate = {2014-01-01},<br \/>\r\nurldate = {2014-01-01},<br \/>\r\nbooktitle = {Proceedings of the 7th NIPS Workshop on Optimization for Machine Learning},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('5','tp_bibtex')\">Close<\/a><\/p><\/div><div class=\"tp_links\" id=\"tp_links_5\" style=\"display:none;\"><div class=\"tp_links_entry\"><ul class=\"tp_pub_list\"><li><i class=\"ai ai-arxiv\"><\/i><a class=\"tp_pub_list\" href=\"https:\/\/arxiv.org\/abs\/1911.12896\" title=\"https:\/\/arxiv.org\/abs\/1911.12896\" target=\"_blank\">https:\/\/arxiv.org\/abs\/1911.12896<\/a><\/li><\/ul><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('5','tp_links')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Adaptive Communication Bounds for Distributed Online Learning\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2014Adaptive.png\" width=\"160\" alt=\"Adaptive Communication Bounds for Distributed Online Learning\" \/><\/div><\/div><h3 class=\"tp_h3\" id=\"tp_h3_2013\">2013<\/h3><div class=\"tp_publication tp_publication_inproceedings\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Kopp, Christine;  Mock, Michael;  Boley, Mario;  May, Michael<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=735\">Privacy-preserving mobility monitoring using sketches of stationary sensor readings<\/a> <span class=\"tp_pub_type tp_  inproceedings\">Proceedings Article<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_in\">In: <\/span><span class=\"tp_pub_additional_booktitle\">Joint European Conference on Machine Learning and Knowledge Discovery in Databases, <\/span><span class=\"tp_pub_additional_pages\">pp. 370\u2013386, <\/span><span class=\"tp_pub_additional_organization\">Springer <\/span><span class=\"tp_pub_additional_year\">2013<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_6\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('6','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_6\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@inproceedings{kamp2013privacy,<br \/>\r\ntitle = {Privacy-preserving mobility monitoring using sketches of stationary sensor readings},<br \/>\r\nauthor = {Michael Kamp and Christine Kopp and Michael Mock and Mario Boley and Michael May},<br \/>\r\nyear  = {2013},<br \/>\r\ndate = {2013-01-01},<br \/>\r\nurldate = {2013-01-01},<br \/>\r\nbooktitle = {Joint European Conference on Machine Learning and Knowledge Discovery in Databases},<br \/>\r\npages = {370--386},<br \/>\r\norganization = {Springer},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {inproceedings}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('6','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Privacy-preserving mobility monitoring using sketches of stationary sensor readings\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Kamp2013Privacypreserving.png\" width=\"160\" alt=\"Privacy-preserving mobility monitoring using sketches of stationary sensor readings\" \/><\/div><\/div><div class=\"tp_publication tp_publication_workshop\"><div class=\"tp_pub_info\"><p class=\"tp_pub_author\"> Kamp, Michael;  Boley, Mario;  G\u00e4rtner, Thomas<\/p><p class=\"tp_pub_title\"><a href=\"https:\/\/michaelkamp.org\/?p=733\">Beating Human Analysts in Nowcasting Corporate Earnings by Using Publicly Available Stock Price and Correlation Features<\/a> <span class=\"tp_pub_type tp_  workshop\">Workshop<\/span> <\/p><p class=\"tp_pub_additional\"><span class=\"tp_pub_additional_booktitle\">2013 IEEE 13th International Conference on Data Mining Workshops, <\/span><span class=\"tp_pub_additional_organization\">IEEE <\/span><span class=\"tp_pub_additional_year\">2013<\/span>.<\/p><p class=\"tp_pub_menu\"><span class=\"tp_bibtex_link\"><a id=\"tp_bibtex_sh_7\" class=\"tp_show\" onclick=\"teachpress_pub_showhide('7','tp_bibtex')\" title=\"Show BibTeX entry\" style=\"cursor:pointer;\">BibTeX<\/a><\/span> | <span class=\"tp_pub_tags_label\">Tags: <\/span><\/p><div class=\"tp_bibtex\" id=\"tp_bibtex_7\" style=\"display:none;\"><div class=\"tp_bibtex_entry\"><pre>@workshop{kamp2013beating,<br \/>\r\ntitle = {Beating Human Analysts in Nowcasting Corporate Earnings by Using Publicly Available Stock Price and Correlation Features},<br \/>\r\nauthor = {Michael Kamp and Mario Boley and Thomas G\u00e4rtner},<br \/>\r\nyear  = {2013},<br \/>\r\ndate = {2013-01-01},<br \/>\r\nurldate = {2013-01-01},<br \/>\r\nbooktitle = {2013 IEEE 13th International Conference on Data Mining Workshops},<br \/>\r\npages = {384--390},<br \/>\r\norganization = {IEEE},<br \/>\r\nkeywords = {},<br \/>\r\npubstate = {published},<br \/>\r\ntppubtype = {workshop}<br \/>\r\n}<br \/>\r\n<\/pre><\/div><p class=\"tp_close_menu\"><a class=\"tp_close\" onclick=\"teachpress_pub_showhide('7','tp_bibtex')\">Close<\/a><\/p><\/div><\/div><div class=\"tp_pub_image_right\"><img decoding=\"async\" name=\"Beating Human Analysts in Nowcasting Corporate Earnings by Using Publicly Available Stock Price and Correlation Features\" src=\"https:\/\/michaelkamp.org\/wp-content\/uploads\/citations\/Gartner2013Beating.png\" width=\"160\" alt=\"Beating Human Analysts in Nowcasting Corporate Earnings by Using Publicly Available Stock Price and Correlation Features\" \/><\/div><\/div><\/div><div class=\"tablenav\"><div class=\"tablenav-pages\"><span class=\"displaying-num\">52 entries<\/span> <a class=\"page-numbers button disabled\">&laquo;<\/a> <a class=\"page-numbers button disabled\">&lsaquo;<\/a> 1 of 2 <a href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;limit=2&amp;tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=&amp;tsr=#tppubs\" title=\"next page\" class=\"page-numbers button\">&rsaquo;<\/a> <a href=\"https:\/\/michaelkamp.org\/?page_id=37&amp;limit=2&amp;tgid=&amp;yr=&amp;type=&amp;usr=&amp;auth=&amp;tsr=#tppubs\" title=\"last page\" class=\"page-numbers button\">&raquo;<\/a> <\/div><\/div><\/div>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-37","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/pages\/37","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/michaelkamp.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=37"}],"version-history":[{"count":17,"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/pages\/37\/revisions"}],"predecessor-version":[{"id":688,"href":"https:\/\/michaelkamp.org\/index.php?rest_route=\/wp\/v2\/pages\/37\/revisions\/688"}],"wp:attachment":[{"href":"https:\/\/michaelkamp.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=37"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}