一种很重要的非监督降维方法，是流形学习算法Laplacian Eigenmap 的线性化方法，在人脸识别中效果非常好。,A very important method of unsupervised dimensionality reduction, manifold learning algorithm is Laplacian Eigenmap linearization method is very effective in face recognition.

流形学习算法lle的线性化方法，是一种非监督的降维方法，比lle的优势在于可以将新的样本点映射到低维空间。-Lle manifold learning algorithm of the linearization method, is a non-supervised dimensionality reduction method has the advantage of being more than lle can sample the new point is mapped to the