SVM parameters optimization algorithm and its application

Abstract

Support vector machines (SVM) is a powerful supervised learning method. It has been used mostly for regression and classification. Some SVM parameters are usually selected artificially, which hampers the efficiency of the SVM algorithm in practical applications. A improved artificial fish swarm algorithm (IAFSA) based on the predatory search strategy of animals was used to optimize the parameters in SVM in this paper. It can effectively avoid the premature convergence and guarantees the variety of solution. With the parameters optimized by IAFSA, the total capability of the SVM was improved. The fault prediction of ship heading angle shows that the SVM optimized by IAFSA can give higher fitting and prediction accuracy than the SVM optimized by basic AFSA.