{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## 第 1 章 绪论\n", "\n", "### 1.1 引言\n", "\n", "机器学习所研究的主要内容，是关于在计算机上从数据中产生 **模型**（model）的算法，即 **学习算法**（learning algorithm）。\n", "\n", "### 1.2 基本术语\n", "\n", "要进行机器学习，先要有数据，比如：\n", "\n", "|色泽|根蒂|敲声|\n", "|---|----|---|\n", "|青绿|蜷缩|浊响|\n", "|乌黑|稍蜷|沉闷|\n", "|浅白|硬挺|清脆|\n", "\n", "这组记录的集合称为 **数据集**（data set），其中的每条记录称为一个 **示例**（instance）或 **样本**（sample）。\n", "\n", "### 1.3 假设空间\n", "\n", "### 1.4 归纳偏好\n", "\n", "### 1.5 发展历程\n", "\n", "### 1.6 应用现状\n", "\n", "### 1.7 阅读材料\n", "\n", "## 第 2 章 模型评估与选择\n", "\n", "### 2.1 经验误差与过拟合\n", "\n", "### 2.2 评估方法\n", "\n", "### 2.3 性能度量\n", "\n", "### 2.4 比较检验\n", "\n", "### 2.5 偏差与方差\n", "\n", "### 2.6 阅读材料\n", "\n", "## 第 3 章 线性模型\n", "\n", "### 3.1 基本形式\n", "\n", "### 3.2 线性回归\n", "\n", "### 3.3 对数几率回归\n", "\n", "### 3.4 线性判别分析\n", "\n", "### 3.5 多分类学习\n", "\n", "### 3.6 类别不平衡问题\n", "\n", "### 3.7 阅读材料\n", "\n", "## 第 4 章 决策树\n", "\n", "### 4.1 基本流程\n", "\n", "### 4.2 划分选择\n", "\n", "### 4.3 剪枝处理\n", "\n", "### 4.4 连续与缺失值\n", "\n", "### 4.5 多变量决策树\n", "\n", "### 4.6 阅读材料\n", "\n", "## 第 5 章 神经网络\n", "\n", "### 5.1 神经元模型\n", "\n", "### 5.2 感知机与多层网络\n", "\n", "### 5.3 误差拟传播算法\n", "\n", "### 5.4 全局最小与局部最小\n", "\n", "### 5.5 其他常见神经网络\n", "\n", "### 5.6 阅读材料\n", "\n", "## 第 6 章 支持向量机\n", "\n", "### 6.1 间隔与支持向量\n", "\n", "### 6.2 对偶问题\n", "\n", "### 6.3 核函数\n", "\n", "### 6.4 软间隔与正则化\n", "\n", "### 6.5 支持向量回归\n", "\n", "### 6.6 核方法\n", "\n", "### 6.7 阅读材料\n", "\n", "## 第 7 章 贝叶斯分类器\n", "\n", "### 7.1 贝叶斯决策论\n", "\n", "### 7.2 极大似然估计\n", "\n", "### 7.3 朴素贝叶斯分类器\n", "\n", "### 7.4 半朴素贝叶斯分类器\n", "\n", "### 7.5 贝叶斯网\n", "\n", "### 7.6 EM 算法\n", "\n", "### 7.7 阅读材料\n", "\n", "## 第 8 章 集成学习\n", "\n", "### 8.1 个体与集成\n", "\n", "### 8.2 Boosting\n", "\n", "### 8.3 Bagging 与随机森林\n", "\n", "### 8.4 结合策略\n", "\n", "### 8.5 多样性\n", "\n", "### 8.6 阅读材料\n", "\n", "## 第 9 章 聚类\n", "\n", "### 9.1 聚类任务\n", "\n", "### 9.2 性能度量\n", "\n", "### 9.3 距离计算\n", "\n", "### 9.4 原型聚类\n", "\n", "### 9.5 密度聚类\n", "\n", "### 9.6 层次聚类\n", "\n", "### 9.7 阅读材料\n", "\n", "## 第 10 章 降维与度量学习\n", "\n", "### 10.1 k近邻学习\n", "\n", "### 10.2 低维嵌入\n", "\n", "### 10.3 主成分分析\n", "\n", "### 10.4 核化线性降维\n", "\n", "### 10.5 流形学习\n", "\n", "### 10.6 度量学习\n", "\n", "### 10.7 阅读材料\n", "\n", "## 第 11 章 特征选择与稀疏学习\n", "\n", "### 11.1 子集搜索与评价\n", "\n", "### 11.2 过滤式选择\n", "\n", "### 11.3 包裹式选择\n", "\n", "### 11.4 嵌入式选择与 $L_1$ 正则化\n", "\n", "### 11.5 稀疏表示与字典学习\n", "\n", "### 11.6 压缩感知\n", "\n", "### 11.7 阅读材料\n", "\n", "## 第 12 章 计算学习理论\n", "\n", "### 12.1 基础知识\n", "\n", "### 12.2 PAC 学习\n", "\n", "### 12.3 有限假设空间\n", "\n", "### 12.4 VC 维\n", "\n", "### 12.5 Rademacher 复杂度\n", "\n", "### 12.6 稳定性\n", "\n", "### 12.7 阅读材料\n", "\n", "## 第 13 章 半监督学习\n", "\n", "### 13.1 未标记样本\n", "\n", "### 13.2 生成式方法\n", "\n", "### 13.3 半监督 SVM\n", "\n", "### 13.4 图半监督学习\n", "\n", "### 13.5 基于分歧的方法\n", "\n", "### 13.6 半监督聚类\n", "\n", "### 13.7 阅读材料\n", "\n", "## 第 14 章 概率图模型\n", "\n", "### 14.1 隐马尔可夫模型\n", "\n", "### 14.2 马尔可夫随机场\n", "\n", "### 14.3 条件随机场\n", "\n", "### 14.4 学习与推断\n", "\n", "### 14.5 近似推断\n", "\n", "### 14.6 话题模型\n", "\n", "### 14.7 阅读材料\n", "\n", "## 第 15 章 规则学习\n", "\n", "### 15.1 基本概念\n", "\n", "### 15.2 序贯覆盖\n", "\n", "### 15.3 剪枝优化\n", "\n", "### 15.4 一阶规则学习\n", "\n", "### 15.5 归纳逻辑程序设计\n", "\n", "### 15.6 阅读材料\n", "\n", "## 第 16 章 强化学习\n", "\n", "### 16.1 任务与奖赏\n", "\n", "### 16.2 K-摇臂赌博机\n", "\n", "### 16.3 有模型学习\n", "\n", "### 16.4 免模型学习\n", "\n", "### 16.5 值函数近似\n", "\n", "### 16.6 模仿学习\n", "\n", "### 16.7 阅读材料\n", "\n", "## 附录\n", "\n", "### 矩阵\n", "\n", "### 优化\n", "\n", "### 概率分布" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.1" } }, "nbformat": 4, "nbformat_minor": 2 }