基于Hopfield网络的复杂网络社团提取  

Based on discrete Hopfield network is a complex network of associations

作　　者：代婷婷[1] 梅端[2] 

机构地区：[1]昭通学院数学与统计学院,云南昭通657000 [2]广东海洋大学理学院,广东湛江524088

出　　处：《昭通学院学报》2016年第5期19-25,共7页Journal of Zhaotong University

摘　　要：针对复杂网络中的社团提取问题,提出了一种基于离散Hopfield神经网络的社团结构提取算法,该算法的思想为:首先,对复杂网络数据的进行处理;其次,结合社团提取准则模块度函数Q的形式设计Hopfield神经网络的权值向量W和阈值T;最后,根据网络稳定点的输出值提取出网络中的社团结构.仿真实验表明,本文中的Hopfield神经网络算法比谱分算法得出的Q值更优,对网络的划分更接近实际网络.针对复杂网络中的社团提取问题,提出了一种基于离散Hopfield神经网络的社团结构提取算法,该算法的思想为:首先,对复杂网络数据的进行处理;其次,结合社团提取准则模块度函数Q的形式设计Hopfield神经网络的权值向量W和阈值T;最后,根据网络稳定点的输出值提取出网络中的社团结构.仿真实验表明,文中的离散Hopfield神经网络算法比谱分算法得出的Q值更优。For community extraction problem in complex networks, this paper proposes a community structure ex- traction algorithm based on discrete Hopfield neural network, the algorithm ideas as follows., first, to deal with complex network data; Secondly, combining with community extraction criterion module function design Hopfield neural network, in the form of weight vector and the threshold valuer Finally, according to the output value of the stable point to extract the network of community structure. Simulation results show that in this paper, a discrete Hopfield neural network algo- rithm is better than it is concluded that the value of the spectrum points algorithm, the division of the network closer to the actual network.

关 键 词：复杂网络 社团提取 HOPFIELD神经网络 

分 类 号：O157.5[理学—数学] TP301.6[理学—基础数学]