Distributed Personalized Empirical Risk Minimization
This addresses the problem of data heterogeneity in distributed learning for clients with varying resources, representing a new paradigm rather than an incremental improvement.
The paper tackles learning from heterogeneous data sources by proposing Personalized Empirical Risk Minimization (PERM), which learns distinct models for each client through model shuffling, achieving optimal statistical accuracy and overcoming data heterogeneity issues.
This paper advocates a new paradigm Personalized Empirical Risk Minimization (PERM) to facilitate learning from heterogeneous data sources without imposing stringent constraints on computational resources shared by participating devices. In PERM, we aim to learn a distinct model for each client by learning who to learn with and personalizing the aggregation of local empirical losses by effectively estimating the statistical discrepancy among data distributions, which entails optimal statistical accuracy for all local distributions and overcomes the data heterogeneity issue. To learn personalized models at scale, we propose a distributed algorithm that replaces the standard model averaging with model shuffling to simultaneously optimize PERM objectives for all devices. This also allows us to learn distinct model architectures (e.g., neural networks with different numbers of parameters) for different clients, thus confining underlying memory and compute resources of individual clients. We rigorously analyze the convergence of the proposed algorithm and conduct experiments that corroborate the effectiveness of the proposed paradigm.