Parameter selection for SVM using niches genetic algorithm in determined range

An automatic selection method for Support Vector Machine (SVM) parameters using niches genetic algorithm in determined range is proposed. Firstly, the limit range of each parameter is determined by analyzing relation between SVM parameters and model capability. Then regard the determined range as search region and use the niches genetic algorithm to automatically search the optimal parameter values of SVM. The experiments are given to illustrate the effectiveness of the method.