基于密度检测和信息增益的半监督kmeans算法

《基于密度检测和信息增益的半监督kmeans算法》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、目录目录第1章绪论·························································································11.1选题背景及意义················································································11.2国内外研究概况和发展趋势·································································21.3主要研究内容及组织结构····

2、································································3第2章聚类分析理论基础·········································································52.1聚类相关理论····················································································52.1.1数据间的相似性度量·······························

3、·······································52.1.2聚类方法的分类············································································62.2半监督聚类相关理论···········································································82.2.1半监督聚类概述····························································

4、················82.2.2半监督聚类算法的分类··································································92.3常见半监督k-measn算法·····································································92.3.1基于约束的监督信息····································································102.3.2Seeded-kmeans算

5、法······································································112.3.3COP-kmeans算法········································································122.4本章小结·························································································13第3章改进的Seeded-kmeans算法······

6、······················································143.1Seeded-kmeans算法存在的问题····························································143.2基于信息增益的Seeded-kmeans改进算法···············································153.2.1数据集的规一化····························································

7、·············153.2.2信息增益的权值计算方法·····························································153.2.3算法描述···················································································173.2.4实验及分析····································································