Expected path length on random manifolds
Machine Learning
2019-08-21 v1 Machine Learning
Abstract
Manifold learning seeks a low dimensional representation that faithfully captures the essence of data. Current methods can successfully learn such representations, but do not provide a meaningful set of operations that are associated with the representation. Working towards operational representation learning, we endow the latent space of a large class of generative models with a random Riemannian metric, which provides us with elementary operators. As computational tools are unavailable for random Riemannian manifolds, we study deterministic approximations and derive tight error bounds on expected distances.
Cite
@article{arxiv.1908.07377,
title = {Expected path length on random manifolds},
author = {David Eklund and Søren Hauberg},
journal= {arXiv preprint arXiv:1908.07377},
year = {2019}
}