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：
提出了一种基于K-means聚类算法的复杂网络社团结构划分方法。算法基于Fortunato等人提出的边的信息中心度，定义了节点的关联度，并通过节点关联度矩阵来进行聚类中心的选择和节点聚类，从而将复杂网络划分成k个社团，然后通过模块度来确定网络理想的社团结构。该算法有效地避免了K-means聚类算法对初始化选值敏感性的问题。通过Zachary Karate Club和College Football Network两个经典模型验证了该算法的可行性。
英文摘要：
This paper proposed a new detecting method based on K-means cluster algorithm. Through the definition of node link based on information centrality which Fortunato proposed and the selection of the clustering center and the clustering of the node according node link, the approach identified the network to k communities, then identified the ideally community structure according modularity. The algorithm could find clustering center better and it is robust to initialization, so the quality of detecting was improved greatly. It tested the algorithm on the two network data named Zachary Karate Club and College Football Network.
同期刊论文项目
同项目期刊论文
期刊信息
位置：