Manifold learning in metric spaces

Liane Xu, Amit Singer

Laplacian-based methods are popular for the dimensionality reduction of data lying in ${\mathbb{R}}^N$. Several theoretical results for these algorithms depend on the fact that the E…