© 2026 LIN TIANYU. All rights reserved. Preview Version.

Powered bynextjs-blog

GitHub LinkedIn

ArticlePDF机器学习

错误有界学习（Mistake-Bounded Learning）

PublishedJuly 24, 2026Document7 pages

Contents

错误有界模型 (Mistake-Bounded Model)
直觉动机 (Informal Motivation)
形式定义 (Formal Definition)
学习单调析取式 (Learning Monotone Disjunctions)
问题设定 (Setup)
重要区分：什么可见，什么未知 (Visible vs. Unknown)
算法 (The Algorithm)
为什么只会犯"假阳性"错误 (Why Only False Positives)
错误上界的证明 (Proof of Mistake Bound)
通过特征扩展学习一般析取式 (Learning General Disjunctions via Feature Expansion)
问题 (The Problem)
特征扩展技术 (Feature Expansion Technique)
为什么有效 (Why This Works)
算法与结果 (The Algorithm and Result)

1

2

3

4

5

6

7

Thanks for reading.