Shadi Albarqouni, PhD
Deep Learning success comes at the cost of collecting and processing a massive amount of data , which often are not accessible due to privacy issues .
Federated Learning has been recently introduced to allow training DL models without sharing the data .
Taken from Rieke, N., Hancox, J., Li, W., Milletari, F., Roth, H.R., Albarqouni, S., Bakas, S., Galtier, M.N., Landman, B.A., Maier-Hein, K. and Ourselin, S., 2020. The future of digital health with federated learning. NPJ digital medicine, 3(1), pp.1-7.
The principal challenges , to overcome, concern the nature of medical data, namely
Objectives: Learn through read, understand, present, and discuss many scientific papers1 tackling the challenges present in Federated Learning.
1Our pool of papers includes the ones published in NeurIPS, ICML, ICLR, IEEE TMI, MedIA, MICCAI, MIDL, and ISBI.
|16.04.2021||Federated Learning; Challenges, Methods, and Future|
|30.04.2021||Data Heterogeneity I|
|14.05.2021||Data Heterogeneity II|
|25.06.2021||Explainability and Accountability|
If you are interested in this seminsr course, please write a brief motivation paragraph (few lines) showing your interest and your background in Machine/Deep Learning and send it with a subject “FLH_Motivation”, to Shadi Albarqouni (email@example.com). Deadline is 16.02.2021.
Don’t forget to register at TUM matching system 11.02 to 16.02.2021: register via matching.in.tum.de