FeDMRA: Federated Incremental Learning with Dynamic Memory Replay Allocation
This addresses incremental learning challenges in federated healthcare systems, offering a balanced solution for privacy-preserving model adaptation, though it is incremental in nature.
The paper tackled the problem of catastrophic forgetting in federated class-incremental learning with non-IID data by proposing a dynamic memory allocation strategy for exemplar storage, resulting in significant performance improvements on three medical image datasets compared to baselines.
In federated healthcare systems, Federated Class-Incremental Learning (FCIL) has emerged as a key paradigm, enabling continuous adaptive model learning among distributed clients while safeguarding data privacy. However, in practical applications, data across agent nodes within the distributed framework often exhibits non-independent and identically distributed (non-IID) characteristics, rendering traditional continual learning methods inapplicable. To address these challenges, this paper covers more comprehensive incremental task scenarios and proposes a dynamic memory allocation strategy for exemplar storage based on the data replay mechanism. This strategy fully taps into the inherent potential of data heterogeneity, while taking into account the performance fairness of all participating clients, thereby establishing a balanced and adaptive solution to mitigate catastrophic forgetting. Unlike the fixed allocation of client exemplar memory, the proposed scheme emphasizes the rational allocation of limited storage resources among clients to improve model performance. Furthermore, extensive experiments are conducted on three medical image datasets, and the results demonstrate significant performance improvements compared to existing baseline models.