English

On Deep Generative Models for Approximation and Estimation of Distributions on Manifolds

Machine Learning 2023-02-28 v1 Machine Learning

Abstract

Generative networks have experienced great empirical successes in distribution learning. Many existing experiments have demonstrated that generative networks can generate high-dimensional complex data from a low-dimensional easy-to-sample distribution. However, this phenomenon can not be justified by existing theories. The widely held manifold hypothesis speculates that real-world data sets, such as natural images and signals, exhibit low-dimensional geometric structures. In this paper, we take such low-dimensional data structures into consideration by assuming that data distributions are supported on a low-dimensional manifold. We prove statistical guarantees of generative networks under the Wasserstein-1 loss. We show that the Wasserstein-1 loss converges to zero at a fast rate depending on the intrinsic dimension instead of the ambient data dimension. Our theory leverages the low-dimensional geometric structures in data sets and justifies the practical power of generative networks. We require no smoothness assumptions on the data distribution which is desirable in practice.

Keywords

Cite

@article{arxiv.2302.13183,
  title  = {On Deep Generative Models for Approximation and Estimation of Distributions on Manifolds},
  author = {Biraj Dahal and Alex Havrilla and Minshuo Chen and Tuo Zhao and Wenjing Liao},
  journal= {arXiv preprint arXiv:2302.13183},
  year   = {2023}
}
R2 v1 2026-06-28T08:49:37.284Z