FedPID: An Aggregation Method for Federated Learning
This work addresses aggregation challenges in federated learning for medical imaging, specifically tumor segmentation, but is incremental as it builds directly on previous methods like FedCostWAvg and FedPIDAvg.
The paper tackles the problem of aggregating models in federated learning with varying dataset sizes by proposing FedPID, an improved aggregation strategy that adjusts training iterations based on dataset size distribution and modifies the integral computation from previous methods. It resulted in a winning submission to the Federated Tumor Segmentation Challenge 2024, building on prior successes in FETS21 and FETS2022.
This paper presents FedPID, our submission to the Federated Tumor Segmentation Challenge 2024 (FETS24). Inspired by FedCostWAvg and FedPIDAvg, our winning contributions to FETS21 and FETS2022, we propose an improved aggregation strategy for federated and collaborative learning. FedCostWAvg is a method that averages results by considering both the number of training samples in each group and how much the cost function decreased in the last round of training. This is similar to how the derivative part of a PID controller works. In FedPIDAvg, we also included the integral part that was missing. Another challenge we faced were vastly differing dataset sizes at each center. We solved this by assuming the sizes follow a Poisson distribution and adjusting the training iterations for each center accordingly. Essentially, this part of the method controls that outliers that require too much training time are less frequently used. Based on these contributions we now adapted FedPIDAvg by changing how the integral part is computed. Instead of integrating the loss function we measure the global drop in cost since the first round.