核主成分分析KPCA算法，经过核变换将样本映射到线性可分的高维空间，再进行PCA降维。包括训练、测试、识别整个过程-KPCA kernel principal component analysis algorithm through nuclear transformation samples are mapped to linearly separable high-dimensional space, then PCA dimensionality reduction. Including training, testing, recognize the entire process