Evaluating the Effects of Parameters Setting on the Performance of Genetic Algorithm Using Regression Modeling and Sta tistical Analysis

Among various heuristics techniques, genetic algorithm (ga) is one of the most widelyused techniques which has successfully been applied on a variety of complex combinatorialproblems. the performance of ga largely depends on the proper selection of its parametersvalues; including crossover mechanism, probability of crossover, population size and mutationrate and selection percent. in this paper, based on design of experiments (doe) approach andregression modeling, the effects of tuning parameters on the performance of genetic algorithmhave been evaluated. as an example, ga is applied to find a shortest distance for a well-knowntravelling salesman problem with 48 cities. the proposed approach can readily be implementedto any other optimization problem. to develop mathematical models, computationalexperiments have been carried out using a 4-factor 5-level central composite design (ccd)matrix. three types of regression functions models have been fitted to relate ga variables to itsperformance characteristic. then, statistical analyses are performed to determine the best andmost fitted model. analysis of variance (anova) results indicate that the second orderfunction is the best model that can properly represent the relationship between ga importantvariables and its performance measure (solution quality).