这是读好的ORL 和YALE人脸库数据， 用LODA加载后， 变量train 代表是训练样本， test 代码是测试样本。-This is a good time ORL and Yale face database data, using Loda loaded, variable train representatives training samples, test code is to test the samples.