KPCA是一种基于核的主要成分分析，是一种由线性到非线性之间的桥梁。通过非线性函数把输入空间映射到高维空间，在特征空间中间型数据处理，引入核函数，把非线性变换后的特征空间内积运算转换为原始空间的核函数计算。

基本思想是通过某种隐士方法将输入空间映射到某个高维空间（特征空间），并在特征空间实现PCA。对该算法进行了详细的说明-KPCA is a kernel-based principal components analysis, is a bridge between the linear to nonlinear. Nonlinear function to map the input space into a high dimensional space, in the middle of the feature space, data processing, the introduction of kernel function, product operation in the non-linear transformation of feature space for the kernel function of the original space calculation.

The basic idea is that the input space by some kind of hermit method is mapped to a higher dimensional space (feature space), and the PCA in the feature space. The algorithm is a detailed descr iption of