Chaotic Ants Swarm Optimize Least Square Support Vector Machine

Abstract

This paper focuses on intelligent analysis of small samples in geotechnical engineering, the calculation process of SVM is simplified by applying LSSVM, the generalization performance of SVM is maintained utmost by using leave one out method, the approximation performance of SVM is optimized by applying wavelet kernel constructed from Marr wavelet, the parameters of SVM is optimized quickly and comprehensively through applying chaotic ants swarm algorithm, the convergence of chaotic ants swarm algorithm is accelerated by reconstituting part of ant-matrix in the process of optimization, and lastly the effectiveness of chaotic ants swarm optimize LSSVM algorithm is proven through practical engineering.