Poster

Improved Graph Laplacian via Geometric Self-Consistency

Dominique Perrault-Joncas ⋅ Marina Meila ⋅ James McQueen

Keywords: Kernel Methods Semi-Supervised Learning Unsupervised Learning Nonlinear Dimensionality Reduction and Manifold Learning Hyperparameter Selection

2017 Poster

[ Paper]

Abstract

We address the problem of setting the kernel bandwidth, epps, used by Manifold Learning algorithms to construct the graph Laplacian. Exploiting the connection between manifold geometry, represented by the Riemannian metric, and the Laplace-Beltrami operator, we set epps by optimizing the Laplacian's ability to preserve the geometry of the data. Experiments show that this principled approach is effective and robust

Chat is not available.