FeDMRA: Federated Incremental Learning with Dynamic Memory Replay Allocation

arXiv stat.ML / 3/31/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces FeDMRA, a federated class-incremental learning approach tailored to federated healthcare settings where client data is non-IID and traditional continual learning assumptions fail.
It proposes a dynamic memory replay allocation strategy that decides how to distribute a limited exemplar buffer across clients rather than using a fixed allocation scheme.
By accounting for data heterogeneity and explicitly targeting performance fairness across clients, FeDMRA aims to reduce catastrophic forgetting while maintaining balanced learning.
The method is evaluated on three medical image datasets, where experiments reportedly show significant performance gains over existing baseline models.

Abstract

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.