对最小二乘支持向量机的核函数进行参数优化，最后得到分类更加准确的分类器-For least squares support vector machine (SVM) kernel function parameter optimization, finally get a more accurate classification of classifier