欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了分析复杂网络和复杂系统的结构和功能特性,提出一种基于谱聚类和主成分分析（principle component analysis,PCA）的网络社团结构检测算法.利用主成分分析方法分析网络中社团结构的拓扑特点,通过压缩网络数据获得网络主要信息,提出了用于确定传统谱聚类中特征向量个数的方法,并在此基础上改进了谱聚类算法.该算法应用于海豚网络和足球网络等网络实例.实验结果表明,该算法可以根据网络结构动态获得特征向量个数,社团划分结果可行有效.
英文摘要：
To reveal and analyze the structural and functional properties in complex networks and systems,a method based on spectral analysis and principle component analysis is proposed to detect community structure.By analyzing the topology of network with the principle component analysis,the major information of the network is extracted from adjacency matrix of network.The Major information decides the correct number of eigenvectors in spectral method and then the improved spectral method based on PCA is established.Finally,the proposed method is applied to the dolphin social network and the football network,and the results demonstrate the performances of the method.
同期刊论文项目
同项目期刊论文
期刊信息
位置：