Wasserstein Bounds for generative diffusion models with Gaussian tail targets

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Xixian, Wang, Zhongjian
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909821859528704
author Wang, Xixian
Wang, Zhongjian
author_facet Wang, Xixian
Wang, Zhongjian
contents We present an estimate of the Wasserstein distance between the data distribution and the generation of score-based generative models. The sampling complexity with respect to dimension is $\mathcal{O}(\sqrt{d})$, with a logarithmic constant. In the analysis, we assume a Gaussian-type tail behavior of the data distribution and an $ε$-accurate approximation of the score. Such a Gaussian tail assumption is general, as it accommodates a practical target - the distribution from early stopping techniques with bounded support. The crux of the analysis lies in the global Lipschitz bound of the score, which is shown from the Gaussian tail assumption by a dimension-independent estimate of the heat kernel. Consequently, our complexity bound scales linearly (up to a logarithmic constant) with the square root of the trace of the covariance operator, which relates to the invariant distribution of the forward process.
format Preprint
id arxiv_https___arxiv_org_abs_2412_11251
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Wasserstein Bounds for generative diffusion models with Gaussian tail targets
Wang, Xixian
Wang, Zhongjian
Machine Learning
Numerical Analysis
Analysis of PDEs
We present an estimate of the Wasserstein distance between the data distribution and the generation of score-based generative models. The sampling complexity with respect to dimension is $\mathcal{O}(\sqrt{d})$, with a logarithmic constant. In the analysis, we assume a Gaussian-type tail behavior of the data distribution and an $ε$-accurate approximation of the score. Such a Gaussian tail assumption is general, as it accommodates a practical target - the distribution from early stopping techniques with bounded support. The crux of the analysis lies in the global Lipschitz bound of the score, which is shown from the Gaussian tail assumption by a dimension-independent estimate of the heat kernel. Consequently, our complexity bound scales linearly (up to a logarithmic constant) with the square root of the trace of the covariance operator, which relates to the invariant distribution of the forward process.
title Wasserstein Bounds for generative diffusion models with Gaussian tail targets
topic Machine Learning
Numerical Analysis
Analysis of PDEs
url https://arxiv.org/abs/2412.11251