文章透彻解读聚类分析及案例实操.docxVIP

文章透彻解读聚类分析及案例实操

目录

一、聚类分析概述............................................3

1.聚类分析定义..........................................4

1.1聚类分析是一种无监督学习方法.......................4

1.2目的是将相似的对象组合在一起.......................5

2.聚类分析分类..........................................6

2.1根据数据类型分为数值聚类和类别聚类.................7

2.2根据目标函数分为划分聚类和层次聚类.................9

二、聚类分析理论基础.......................................10

1.距离度量方法.........................................11

1.1欧氏距离..........................................13

1.2曼哈顿距离........................................14

1.3余弦相似度........................................15

1.4皮尔逊相关系数....................................16

2.聚类有效性指标.......................................17

三、聚类分析算法...........................................18

1.K-均值聚类...........................................19

1.1算法原理..........................................21

1.2算法步骤..........................................22

1.3收敛条件和异常值处理..............................24

2.层次聚类.............................................25

2.1算法原理..........................................26

2.2算法步骤..........................................27

2.3凝聚度量和链接度量................................28

四、案例实操...............................................30

1.客户分群.............................................31

1.1数据准备..........................................33

1.2聚类结果分析......................................34

1.3结果应用..........................................35

2.商品推荐.............................................36

2.1数据准备..........................................37

2.2聚类结果分析......................................38

2.3结果应用..........................................39

3.新闻分类.............................................40

3.1数据准备..........................................41

3.2聚类结果分析......................................42

3.3结果应用..........................................44

五、聚类分析应用场景........