Adaptive Weight Aggregation in Federated Learning for Brain Tumor Segmentation
Khan, Muhammad Irfan; Jafaritadi, Mojtaba; Alhoniemi, Esa; Kontio, Elina; Khan, Suleiman A. (2022)
Khan, Muhammad Irfan
Jafaritadi, Mojtaba
Alhoniemi, Esa
Kontio, Elina
Khan, Suleiman A.
Editoija
Crimi, Alessandro
Bakas, Spyridon
Springer
2022
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2022092660132
https://urn.fi/URN:NBN:fi-fe2022092660132
Tiivistelmä
We introduce similarity weighted aggregation, a principled and efficient method for regularized weight aggregation in federated learning. Our method is adapted to non-IID collaborators and is simultaneously cost-efficient. This is the first method to propose a slidingwindow to select the collaborators, to the best of our knowledge. We demonstrate our method on the federate training task of the FeTS 2021 challenge.We proposed two variations coined SimilarityWeighted Aggregation (SimAgg) and Regularized Aggregation (RegAgg). SimAgg results on internal validation data demonstrate that the proposed method outperforms the baseline FedAvg. The method SimAgg by our team HTTUAS won 2nd position on both leaderboards in FeTS2021 challenge. SimAgg is the only method to be among the top performing methods on both the leaderboards, making it robust and reliable to data variations. Our solution is open sourced at: https://github.com/dskhanirfan/ FeTS2021