机器学习实验手册

　机器学习实验手册

　目录 目录........................................................................................................................................... 2 第一章 软件包的下载、安装及配置 ..................................................................................... 7 1.1 软件包的下载准备 .................................................................................................... 7 1.1.1 Anaconda 下载 ........................................................................................... 7 1.1.2 PyCharm 编辑器下载................................................................................. 11 1.2 windows 下安装步骤 ............................................................................................ 12 1.2.1 安装 python 科学计算 Anaconda .......................................................... 12 1.2.2 验证环境配置 .............................................................................................. 17 1.2.3 PyCharm 编辑器的安装 ............................................................................. 17 1.2.4 PyCharm 配置 ............................................................................................ 21 1.3 Mac OS X 下安装步骤 .......................................................................................... 23 1.3.1 安装 python 科学计算 Anaconda .......................................................... 23 1.3.2 验证环境配置 .............................................................................................. 28 1.3.3 PyCharm 编辑器的安装及配置 ................................................................. 29 1.4 Linux 下安装步骤（以 ubuntu 系统为例）

　....................................................... 32 1.4.1 安装 python 科学计算 Anaconda .......................................................... 32 1.4.2 验证环境配置 .............................................................................................. 35 1.4.3 PyCharm 编辑器的安装及配置 ................................................................. 35 第二章 回归 ........................................................................................................................... 40 2.1 线性回归 ................................................................................................................. 40

　 2.1.1 实验数据 ...................................................................................................... 40 2.1.2 实验过程: ..................................................................................................... 42 2.1.3 结果分析：

　................................................................................................... 49 2.1.4 注意事项：

　................................................................................................... 50 2.2 Logistic 回归 .......................................................................................................... 51 2.2.1 实验数据: ..................................................................................................... 51 2.2.2 实验过程: ..................................................................................................... 52 2.2.3 结果分析：

　.................................................................................................. 54 2.2.4 注意事项: ........................................................................ 错误!未定义书签。

　第三章 决策树和随机森林 ................................................................................................... 56 3.1 实验数据 ................................................................................................................... 56 3.2 特性和思想 ............................................................................................................... 56 3.3 实验过程 ................................................................................................................... 59 3.3.1 决策树实现 .................................................................................................... 59 3.3.2 Matplotlib 绘制树形图 ..................................................... 错误!未定义书签。

　3.3.3 使用决策树分类 ............................................................... 错误!未定义书签。

　3.3.4 使用随机生成的数据集构造可视化决策树 ................... 错误!未定义书签。

　3.3.5 通过真实汽车数据构造实际可视化决策树 ................... 错误!未定义书签。

　3.3.6 检验生成的决策树性能 ................................................... 错误!未定义书签。

　3.4 实验结果 ...................................................................................... 错误!未定义书签。

　3.4.1 初步结果 ........................................................................... 错误!未定义书签。

　3.4.2 改进结果 ........................................................................... 错误!未定义书签。

　 3.5 实验总结 ...................................................................................... 错误!未定义书签。

　第四章 支持向量机 SVM ..................................................................................................... 67 4.1 实验数据 ................................................................................................................. 67 4.2 LIBSVM 简介 ............................................................................................................. 72 4.3 LIBSVM 在当前实验数据的调用过程 ...................................................................... 72 4.4 实验效果分析 .......................................................................................................... 77 第五章 聚类 ........................................................................................................................... 78 5.1 实验概要 .................................................................................................................. 78 5.2 实验输入描述 .......................................................................................................... 78 5.3 实验步骤 ................................................................................................................. 79 5.4 评价标准 ................................................................................................................. 80 5.4.1 Adjusted Rand Index（ARI）

　................................................................ 80 5.4.2 Homogeneity（同质性）

　......................................................................... 81 5.4.3 Completeness（完整性）

　........................................................................ 81 5.5 实验结果及分析 ..................................................................................................... 81 第六章 EM 算法 .................................................................................................................... 82 6.1 实验概要 .................................................................................................................. 82 6.2 实验输入描述 .......................................................................................................... 82 6.3 实验步骤 ................................................................................................................. 83 6.4 实验结果及分析 ..................................................................................................... 84 第七章 LDA ........................................................................................................................... 86 7.1 目标任务 .................................................................................................................. 86

　 7.2 实验环境 ................................................................................................................. 86 7.3 实验数据 ................................................................................................................. 86 数据集一：网易新闻 ............................................................................................. 87 数据集二：搜狗新闻 ............................................................................................. 89 7.4 实验设计 ................................................................................................................. 91 7.5 实验过程 ................................................................................................................. 92 7.5.1 数据的输入输出：

　...................................................................................... 92 7.5.2 实验步骤 ...................................................................................................... 93 实验准备：配置环境 ............................................................................................. 93 实验一：测试数据集的主题分布 ......................................................................... 95 实验二：测试 LDA 参数的改变对主题分布的影响 ........................................... 98 7.6 附：Java 版 LDA 的使用和结果 ........................................................................ 100 7.6.1 原始数据：

　................................................................................................ 100 7.6.2 分词处理：

　................................................................................................ 101 7.6.3 LDA 主题分类 ............................................................................................ 102 第八章 隐马尔科夫模型 HMM ......................................................................................... 105 8.1 目标任务：

　........................................................................................................... 105 8.2 实验数据: ............................................................................................................... 105 8.3 实验过程：

　........................................................................................................... 105 8.3.1 实验准备：相关库的安装 ................................................................................ 105 8.3.2 数据的输入输出: ....................................................................................... 117 8.3.3 实验步骤: ................................................................................................... 118

　 8.4 结果分析 ......................................

推荐访问:机器 实验 手册