支持向量回归机的参数选择程序，通过交叉验证选择惩罚参数、损失函数以及核函数-SVM parameter selection procedure

基于支持向量机SVM原理，引入惩罚系数，在libsvm3.14安装包基础上实现了支持向量回归matlab编程(Based on the SVM principle of support vector machine, the penalty coefficient is introduced, and support vector regression matlab programming is implemented on the basis of libsvm3.14 installatio

使用局部二次逼近的方法（LQA），求解带有SCAD惩罚项的线性回归问题。(Use the local quardratic approximation (LQA) method to solve the problem of the smoothly clipped absolute deviation penalty (SCAD) in linear regression.)