复杂网络社团发现的聚类算法分析-计算机软件与理论专业论文.docxVIP

西安理工大学硕士学 西安理工大学硕士学位论文 IV IV i i The community structure is different with different similarity measures and of adjacent degree clustering algorithm. So it is the same key about selecting an appropriate similarity measure and adjacent degree in the community detecting. In this paper, through the comparison of different similarity measures and adjacent degree combined with network itself properties; we find a set of more reasonable choice. This paper puts forward a method of finding the shortest path of undirected network node based on the signal transmission algorithm. This paper improves the original spectral clustering algorithm method, and the experimental results show that this method gets good community structures with the traditional spectral clustering method compared. Key words: complex network; community detecting; signal transmission; clustering; accuracy 目录 目 录 HYPERLINK \l _bookmark0 1 绪 论 1 HYPERLINK \l _bookmark1 1.1 研究背景和意义 1 HYPERLINK \l _bookmark2 1.2 国内外研究动态 2 HYPERLINK \l _bookmark3 1.3 本文研究内容以及创新点 4 HYPERLINK \l _bookmark4 1.4 本文章节安排 5 HYPERLINK \l _bookmark5 2 复杂网络社团结构及其基本性质 7 HYPERLINK \l _bookmark6 2.1 复杂网络的定义及表示方法 7 HYPERLINK \l _bookmark7 2.1.1 复杂网络的定义 7 HYPERLINK \l _bookmark8 2.1.2 复杂网络的表示方法 8 HYPERLINK \l _bookmark9 2.1.3复杂网络的统计特性9 HYPERLINK \l _bookmark10 2.1.4 复杂网络的结构特性 12 HYPERLINK \l _bookmark11 2.2 复杂网络社团发现中常使用的思路 15 HYPERLINK \l _bookmark12 2.2.1 全局方法 15 HYPERLINK \l _bookmark13 2.2.2 局部方法 16 HYPERLINK \l _bookmark14 2.2.3 基于节点相似度的方法 16 HYPERLINK \l _bookmark15 2.3 社团发现的度量函数 18 HYPERLINK \l _bookmark16 2.3.1 模块度检测 18 HYPERLINK \l _bookmark17 2.3.2 社团的准确度、精确度和最佳社团个数 18 HYPERLINK \l _bookmark18 2.3.3 边强度和边比率 19 HYPERLINK \l _bookmark19 2.4 本章总结 20 HYPERLINK \l _bookmark20 3 基于聚类的社团发现算法研究 21 HYPERLINK \l _bookmark21 3.1 聚类的过程 21 HYPERLINK \l _bookmark22 3.2 聚类算法的分类概述以及要求 24 HYPERLINK \l _bookmark23 3.2.1 聚类算法分类 24 HYPERLINK \l _bookmark24 3.2.2 聚类算法的要求 26 HYPERLINK \l _bookmar