Guardado en:
Detalles Bibliográficos
Autores principales: Hosseini, Hesam, Cao, Ying, Sayed, Ali H.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2509.19234
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912601693224960
author Hosseini, Hesam
Cao, Ying
Sayed, Ali H.
author_facet Hosseini, Hesam
Cao, Ying
Sayed, Ali H.
contents Algorithmic stability is an established tool for analyzing generalization. While adversarial training enhances model robustness, it often suffers from robust overfitting and an enlarged generalization gap. Although recent work has established the convergence of adversarial training in decentralized networks, its generalization properties remain unexplored. This work presents a stability-based generalization analysis of adversarial training under the diffusion strategy for convex losses. We derive a bound showing that the generalization error grows with both the adversarial perturbation strength and the number of training steps, a finding consistent with single-agent case but novel for decentralized settings. Numerical experiments on logistic regression validate these theoretical predictions.
format Preprint
id arxiv_https___arxiv_org_abs_2509_19234
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Stability and Generalization of Adversarial Diffusion Training
Hosseini, Hesam
Cao, Ying
Sayed, Ali H.
Machine Learning
Signal Processing
Algorithmic stability is an established tool for analyzing generalization. While adversarial training enhances model robustness, it often suffers from robust overfitting and an enlarged generalization gap. Although recent work has established the convergence of adversarial training in decentralized networks, its generalization properties remain unexplored. This work presents a stability-based generalization analysis of adversarial training under the diffusion strategy for convex losses. We derive a bound showing that the generalization error grows with both the adversarial perturbation strength and the number of training steps, a finding consistent with single-agent case but novel for decentralized settings. Numerical experiments on logistic regression validate these theoretical predictions.
title Stability and Generalization of Adversarial Diffusion Training
topic Machine Learning
Signal Processing
url https://arxiv.org/abs/2509.19234