经典极大团挖掘算法研究报告
首发时间:2015-10-09
摘要:极大团挖掘算法是图挖掘领域的核心算法之一,在社交网络分析、行为与认知网络结构的研究、金融网络的统计分析与动态网络的聚类方面有着广泛的应用。本文以半个多世纪以来七个经典的极大团挖掘算法为研究内容,介绍了各个算法的理论依据与实现方式。在单机环境下以不同稠密程度、不同节点规模的图作为输入,测量各个算法挖掘极大团的时间消耗与空间消耗,一共做了240组,1680个实验。最后得到各个算法的适用范围,给出不同环境下挖掘极大团的建议,并对挖掘大规模的输入图的问题作出展望。实验用到的七个算法分别是Base BK算法,不同启发式选择标志节点的Improved BK系列算法以及Kose算法的RAM版本。
关键词: 算法理论 极大团 Base BK算法 Improved BK算法 Kose算法
For information in English, please click here
A Report on Classic Maximal Clique Enumerating Algorithms
Abstract:Maximal clique enumerating algorithm is one of the core algorithms in graph mining area. It has been widely used in the research of social network analysis, behavior and cognitive network structure, the statistical analysis of financial network and clustering of dynamic network. In this paper, we have studied the seven classical algorithms of maximal clique enumerating in the half century, and the theoretical basis and implementation methods of each algorithm are introduced. In the single machine environment with different density, different node size of the graph as the input, the time consumption and space consumption for each of the algorithms to mine maximal cliques are reported. There are total of 240 groups, 1680 experiments. Finally, the application scope of each algorithm is obtained, and the suggestion of mining maximal group under different conditions is given. The seven algorithms used in the experiment are Base BK algorithm, the series of Improved BK algorithms and RAM version of Kose algorithm.
Keywords: Algorithm Theory Maximal Clique Base BK Algorithm Improved BK Algorithm Kose Algorithm
基金:
论文图表:
引用
No.4655709109521214****
同行评议
共计0人参与
勘误表
经典极大团挖掘算法研究报告
评论
全部评论