On empirical Hodge Laplacians under the manifold hypothesis
By: Jan-Paul Lerch, Martin Wahl
Potential Business Impact:
Improves how computers understand shapes in data.
Given i.i.d. observations uniformly distributed on a closed submanifold of the Euclidean space, we study higher-order generalizations of graph Laplacians, so-called Hodge Laplacians on graphs, as approximations of the Laplace-Beltrami operator on differential forms. Our main result is a high-probability error bound for the associated Dirichlet forms. This bound improves existing Dirichlet form error bounds for graph Laplacians in the context of Laplacian Eigenmaps, and it provides insights into the Betti numbers studied in topological data analysis and the complementing positive part of the spectrum.
Similar Papers
Hodge Laplacians and Hodge Diffusion Maps
Machine Learning (CS)
Finds hidden shapes in complicated data.
Laplace Learning in Wasserstein Space
Machine Learning (CS)
Helps computers learn from messy data better.
Re-imagining Spectral Graph Theory
Combinatorics
New math tool connects different kinds of networks.