stat.MLSep 10, 2026

Generalization Analysis of Distributed Kernel-based Robust Gradient Descent Algorithms

Authors: Jun-Yi MengZheng-Chu GuoYuan Mao

Abstract

In this paper, we investigate the generalization performance of distributed gradient descent algorithms in a reproducing kernel Hilbert space under a robust loss function lσl_σ. By exploiting the spectral characterization of gradient descent together with the intrinsic properties of robust loss functions, we establish optimal learning rates for the distributed kernel-based robust gradient descent (DKRGD) algorithm with an appropriately chosen scale parameter σσ. The proposed parameter choice of σσ simultaneously alleviates the saturation phenomenon and guarantees statistical robustness. A key technical contribution is a novel error analysis that provides substantially sharper bounds for products of operators, thereby significantly relaxing existing restrictions on the maximum number of local machines while retaining optimal learning rates. Finally, we develop a communication-efficient strategy that further improves the convergence performance of DKRGD.

Explore similar work

CardsList
  1. Distributionally-Robust Learning to Optimize

    May 7, 2026Vinit Ranjan, Jisun Park, Bartolomeo StellatoRobust OptimizationDistributional Learning