Introduction - If you have any usage issues, please Google them yourself
Laplacian Eigenmap is a kind of nonlinear dimensionality reduction technique which based on manifold study, it choose the weights W using the heat kernel and it can be used for image segmentation to promote clustering performance.