| Since the study of complex networks emerging in the late 20th century, it is rapidly spread to the research of various fields. It has caused many scientific research personnel's widespread attention, which in the domain of mathematics, physics, the computer,the sociology and so on. The problem of community structure detection is an important branch of the study of complex networks, now it becoming a hot spot of field. Genetic Algorithms (GA) was proposed in 1969 by Professor Holland of United States University of Michigan. This article combines the ideas of genetic algorithms with the characteristics of community division in complex network, proposed single parent genetic algorithm for solving this problem.Detection of community structure of complex networks is the basis of the study of community structure. At present, there is not a widely recognized and uniform standard for the concept of community structures and the merits of the results of detection. In this article, firstly we made a presentation for the research background of complex network, and then give a more comprehensive review of the concepts of community structure and some measure of community quality standards of community structure detection. Second, this article describes some representative methods of the detection of community structure in complex network. Finally, combine with some of new research results about community structure in complex networks, we use the Partheno-Genetic Algorithm (PGA) for solving this problem, and than test the algorithm with some classic example. In the test,we verify the validity of the algorithm, analysis and comparison the results of make some current evaluation as a fitness function.Innovative work of this article:(1) Use partheno-genetic algorithm for complex networks on the community division, constructed a new fitness function QF, encoding and mutation operators.(2) Constructed three optimization operators, The result indicate it's availability。(3) Design the method for initialization, which advance the efficiency of the algorithm(4) Programming for calculation and analysis the results. Evaluate several function like Q,D,C,Q... |
