基于Laplace矩阵Jordan型的复杂网络聚类.pdfVIP

第35 卷第3 期 通 信 学 报 Vol.35 No. 3 20 14 年3 月 Journal on Communications March 20 14 doi: 10.3969/j.issn. 1000-436x.2014.03.002 基于Laplace 矩阵Jordan 型的复杂网络聚类算法 牛建伟，戴彬，童超，霍冠英，彭井 （北京航空航天大学虚拟现实技术与系统国家重点实验室，北京 100191） 摘 要：在目前复杂网络聚类算法中，基于 Laplace 特征值的谱聚类方法具有严密的数学理论和较高的精度，但 受限于该方法对簇结构数量、规模等先验知识的依赖，难以实际应用。针对这一问题，基于Laplace 矩阵的Jordan 型变换，提出了一种先验知识的自动获取方法，实现了基于Jordan 矩阵特征向量的初始划分。基于Jordan 型特征 值定义了簇结构的模块化密度函数，并使用该函数和初始划分结果完成了高精度聚类算法。该算法在多个数据集 中的实验结果表明，与目前主流的Fast-Newman 算法、Girvan-Newman 算法相比，基于Laplace 矩阵Jordan 型聚 类算法在不依赖先验知识的情况下，实现了更高的聚类精度，验证了先验知识获取方法的有效性和合理性。 关键词：复杂网络；聚类算法；Laplace 矩阵；Jordan 型；先验知识获取 中图分类号：TP301.6 文献标识码：A 文章编号：1000-436X(2014)03-0011-11 Complex network clustering algorithm based on Jordan-form of Laplace-matrix NIU Jian-wei, DAI Bin, TONG Chao, HUO Guan-ying, PENG Jing (State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191) Abstract: Among existing clustering algorithms, the graph-Laplacian-based spectrum clustering algorithm has rigorous theoretical basis and high accuracy. However, the application of this algorithm is limited by its dependence on the prior knowledge, such as the number and the size of clusters. Based on the Jordan form of graph Laplacian, an algorithm was proposed which can obtain the prior knowledge, and perform the primary clustering based on the eigenvalues of the Jor- dan form. The modularity density function of clusters was defined, and an improved spectrum clustering algorithm with the help of the function and the primary clustering was proposed. The experiments were conducted on diverse