arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.03514cs.LGmath.OCstat.ML
收藏

通用插值学习系统中泛化稳定性的函数空间边界

A Function-Space Stability Boundary for Generalization in Interpolating Learning Systems

Ronald Katende

AI总结:

本文提出了一种函数空间稳定性边界,用于分析通用插值学习系统中稳定性与泛化的关系,揭示了稳定性并非唯一解释泛化的机制。

摘要:

现代学习系统常常在插值训练数据的同时仍能良好泛化,但尚不清楚何时算法稳定性解释了这种行为。我们将训练建模为函数空间轨迹,并测量沿此轨迹的单样本扰动敏感性。我们提出了一种收缩传播条件和通过展开所得递归获得的稳定性证书。小证书表明稳定性导致的泛化,同时我们证明存在插值区域中风险小但无法保持此类收缩敏感性,显示稳定性并非普遍解释。实验表明证书增长可预测不同优化器、步长和数据集扰动下的泛化差异。该框架因此识别了稳定性解释泛化的区域以及必须由其他机制解释成功区域。

英文摘要:

Modern learning systems often interpolate training data while still generalizing well, yet it remains unclear when algorithmic stability explains this behavior. We model training as a function-space trajectory and measure sensitivity to single-sample perturbations along this trajectory. We propose a contractive propagation condition and a stability certificate obtained by unrolling the resulting recursion. A small certificate implies stability-based generalization, while we also prove that there exist interpolating regimes with small risk where such contractive sensitivity cannot hold, showing that stability is not a universal explanation. Experiments confirm that certificate growth predicts generalization differences across optimizers, step sizes, and dataset perturbations. The framework therefore identifies regimes where stability explains generalization and where alternative mechanisms must account for success.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑