PG.zip_混合核函数

SVM faruto version by faruto Email:farutoliyang@gmail.com 2009.11.05 ================================== Content: scaleForSVM:归一化 函数接口: [train_scale,test_scale,ps] = scaleForSVM(train_data,test_data,ymin,ymax) ==================================== pcaForSVM:pca降为预处理 函数接口: [train_pca,test_pca] = pcaForSVM(train,test,threshold) ==================================== fasticaForSVM:ica降为预处理 函数接口: [train_ica,test_ica] = fasticaForSVM(train,test) ==================================== SVMcgForClass:分类问题参数寻找[grid search based on CV] 函数接口: [bestacc,bestc,bestg] = SVMcgForClass(train_label,train,cmin,cmax,gmin,gmax,v,cstep,gstep,accstep) SVMcgForRegress:回归问题参数寻优[grid search based on CV] 函数接口: [mse,bestc,bestg] = SVMcgForRegress(train_label,train,cmin,cmax,gmin,gmax,v,cstep,gstep,msestep) ====================================== psoSVMcgForClass:分类问题参数寻优[pso based on CV] 函数接口: [bestCVaccuarcy,bestc,bestg,pso_option] = psoSVMcgForClass(train_label,train,pso_option) psoSVMcgForRegress:回归问题参数寻优[pso based on CV] 函数接口: [bestCVmse,bestc,bestg,pso_option] = psoSVMcgForRegress(train_label,train,pso_option) ======================================= gaSVMcgForClass:分类问题参数寻优[ga based on CV] 函数接口: [bestCVaccuarcy,bestc,bestg,ga_option] = gaSVMcgForClass(train_label,train,ga_option) gaSVMcgForRegress:回归问题参数寻优[ga based on CV] 函数接口: [bestCVmse,bestc,bestg,ga_option] = gaSVMcgForRegress(train_label,train,ga_option) ======================================