基于AHP-熵权法的复杂网络关键节点识别方法  

Identification of key nodes in a complex network based on AHP-entropy method

作　　者：严栋[1] 张世斌[2] 宗康[1] 胡志华[1] 

机构地区：[1]上海海事大学物流研究中心,上海201306 [2]上海海事大学数学系,上海201306

出　　处：《广西大学学报（自然科学版）》2016年第6期1933-1939,共7页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然科学基金面上项目(11671416);上海市教委科研创新基金资助项目(14YZ115)

摘　　要：复杂网络中的关键节点通常数量较少,但是对网络的影响却很大。现有的关键节点识别方法忽略了网络的整体特性、易受主观因素的影响。根据主观与客观赋权的特点,把熵权法的思想运用到AHP算法,建立了新的复杂网络关键节点识别方法,该方法在一定程度上克服了AHP赋权的主观性以及指标识别能力不足的缺陷。对BA无标度网络和WS小世界网络两个典型的复杂网络模型进行仿真表明,该方法可行有效,能准确识别出网络的关键节点,与现有的关键节点识别方法相比,其识别度更高。The quantity of key nodes in a complex network is usually small,but their impact on the network is crucial. The existing methodsfor identifying key nodes in a complex network are sensitive to subjective factors since they ignore the overall characteristics of the network. According to the features of both subjective and objective weighting,a new method for identifying key nodes in a complex network is proposed by means of combining the idea of the entropy weight method into the AHP algorithm. To some extents,the proposed method overcomes the subjectivity and insufficiency of the indicator identification of AHP. The good performance of the proposed method is verified by simulation for two typical complex network models,the BA scale-free network and the WS small-world network. The proposed method can not only identify key nodes of the network accurately,but also is feasible and effective,but also achieve a higher recognition when comparedwith the existing methods.

关 键 词：复杂网络 关键节点 AHP算法 熵权法 

分 类 号：TP315[自动化与计算机技术—计算机软件与理论]