分析分布式对抗训练的泛化稳定性，发现误差随扰动强度和训练步数增长。
Stability and Generalization of Adversarial Diffusion Training
- 基于稳定性理论分析分布式扩散策略下的对抗训练
- 证明泛化误差随扰动强度和训练步数上升，与单机情况一致
- 适合关注对抗训练泛化能力的算法研究者
算法稳定性是分析泛化性能的经典工具。尽管对抗训练能提升模型鲁棒性，但常面临鲁棒过拟合和泛化差距扩大问题。虽有研究证明了分布式环境下对抗训练的收敛性，其泛化性质仍不明确。本文针对凸损失函数，在扩散策略下开展对抗训练的稳定性-泛化分析，推导出泛化误差随对抗扰动强度和训练步数增长的上界，该结论在单智能体情况下已有支持，但在分布式设置中为首次揭示。逻辑回归上的数值实验验证了理论预测。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。