FedMeS: Personalized Federated Continual Learning Leveraging Local Memory
This work addresses the challenge of training personalized models in federated learning with continual tasks, which is incremental as it builds on existing PFCL methods with a novel memory-based approach.
The paper tackles the problem of Personalized Federated Continual Learning (PFCL), where distributed clients with local task sequences collaborate to train personalized models, and proposes FedMeS to address client drift and catastrophic forgetting, showing it outperforms baselines in average accuracy and forgetting rate across various datasets and client numbers.
We focus on the problem of Personalized Federated Continual Learning (PFCL): a group of distributed clients, each with a sequence of local tasks on arbitrary data distributions, collaborate through a central server to train a personalized model at each client, with the model expected to achieve good performance on all local tasks. We propose a novel PFCL framework called Federated Memory Strengthening FedMeS to address the challenges of client drift and catastrophic forgetting. In FedMeS, each client stores samples from previous tasks using a small amount of local memory, and leverages this information to both 1) calibrate gradient updates in training process; and 2) perform KNN-based Gaussian inference to facilitate personalization. FedMeS is designed to be task-oblivious, such that the same inference process is applied to samples from all tasks to achieve good performance. FedMeS is analyzed theoretically and evaluated experimentally. It is shown to outperform all baselines in average accuracy and forgetting rate, over various combinations of datasets, task distributions, and client numbers.