arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.01276cs.CCcs.DScs.LG
收藏

可采性使学习更容易

Samplability makes learning easier

Stanford(斯坦福大学);MIT(麻省理工学院)

Guy Blanc, Caleb Koch, Jane Lange, Carmen Strassle, Li-Yang Tan

AI总结:

本文研究了可采性PAC学习与标准PAC学习的区别,展示了可采性PAC学习在样本复杂度和计算效率上的优势。

摘要:

标准的PAC学习定义(Valiant 1984)要求学习者在所有分布下成功——即使这些分布难以采样。这与可采性PAC学习(Blum, Furst, Kearns和Lipton 1993)形成对比,后者仅要求学习者在可采样的分布下成功。我们研究这一区别,并展示可采性PAC显著扩展了高效学习者的能力。我们首先构造了一个需要指数样本复杂度的标准PAC概念类,但在可采性PAC中可以以多项式样本复杂度学习。然后我们将这一统计分离提升到计算设置,并在随机 oracle 的背景下获得分离。我们的证明围绕一个新的复杂性原始性——显式逃避集,我们引入并研究了它。这些集合的成员资格易于确定,但极难从其中采样。我们的结果扩展到在线设置,同样展示了当对手被假设为高效而非计算无界时,其景观如何变化。

英文摘要:

The standard definition of PAC learning (Valiant 1984) requires learners to succeed under all distributions -- even ones that are intractable to sample from. This stands in contrast to samplable PAC learning (Blum, Furst, Kearns, and Lipton 1993), where learners only have to succeed under samplable distributions. We study this distinction and show that samplable PAC substantially expands the power of efficient learners. We first construct a concept class that requires exponential sample complexity in standard PAC but is learnable with polynomial sample complexity in samplable PAC. We then lift this statistical separation to the computational setting and obtain a separation relative to a random oracle. Our proofs center around a new complexity primitive, explicit evasive sets, that we introduce and study. These are sets for which membership is easy to determine but are extremely hard to sample from. Our results extend to the online setting to similarly show how its landscape changes when the adversary is assumed to be efficient instead of computationally unbounded.

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。

↑