文章基本信息

标题：Novel Knowledge based Selective Tabu Initialization in Genetic algorithm
本地全文：下载
作者：Manju Sharma ; Sanjay Tyagi
期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2013
卷号：3
期号：5
出版社：S.S. Mishra
摘要：Genetic algorithms facilitate the hybridization of other local search techniques to get the optimal solution and to remove the problem of genetic drift. Success of Genetic Algorithm mainly depends upon the individuals selected in the initial population and the size of population. If the individuals chosen in the initial population are poor, it will result in weaker solutions and premature convergence towards optima. This paper proposes a novel selective tabu initialization method based on tabu search that supplies more fit individuals in the beginning phase itself. The proposed tabu initialization is tested on three different instances from TSPLIB provided by Heidelberg University and the result are compared with simple random initialization and hill climbing based initialization. The implementation has been carried out using MATLAB and result shows that the proposed selective initialization by tabu search outperforms the existing random and hill climbing based initialization scheme used in genetic algorithm in terms of producing more optimal solution and better convergence speed
关键词：Genetic Algorithm; Hill Climbing; Hybrid genetic algorithm; Initialization; Tabu search