FedCostWAvg: A new averaging for better Federated Learning
This work addresses model aggregation in federated learning for medical imaging, specifically tumor segmentation, and is incremental as it builds upon FedAvg.
The paper tackles the problem of aggregating multiple models trained on different datasets in federated learning by proposing a new weight selection method for averaging, which won the MICCAI Federated Tumor Segmentation Challenge 2021 and shows notable improvement in segmentation performance compared to FedAvg.
We propose a simple new aggregation strategy for federated learning that won the MICCAI Federated Tumor Segmentation Challenge 2021 (FETS), the first ever challenge on Federated Learning in the Machine Learning community. Our method addresses the problem of how to aggregate multiple models that were trained on different data sets. Conceptually, we propose a new way to choose the weights when averaging the different models, thereby extending the current state of the art (FedAvg). Empirical validation demonstrates that our approach reaches a notable improvement in segmentation performance compared to FedAvg.