An approximation theory framework for measure-transport sampling algorithms

Author:

Baptista Ricardo,Hosseini Bamdad,Kovachki Nikola,Marzouk Youssef,Sagiv Amir

Abstract

This article presents a general approximation-theoretic framework to analyze measure transport algorithms for probabilistic modeling. A primary motivating application for such algorithms is sampling—a central task in statistical inference and generative modeling. We provide a priori error estimates in the continuum limit, i.e., when the measures (or their densities) are given, but when the transport map is discretized or approximated using a finite-dimensional function space. Our analysis relies on the regularity theory of transport maps and on classical approximation theory for high-dimensional functions. A third element of our analysis, which is of independent interest, is the development of new stability estimates that relate the distance between two maps to the distance (or divergence) between the pushforward measures they define. We present a series of applications of our framework, where quantitative convergence rates are obtained for practical problems using Wasserstein metrics, maximum mean discrepancy, and Kullback–Leibler divergence. Specialized rates for approximations of the popular triangular Knöthe–Rosenblatt maps are obtained, followed by numerical experiments that demonstrate and extend our theory.

Funder

U.S. Department of Energy

Publisher

American Mathematical Society (AMS)

Reference104 articles.

1. Lectures in Mathematics ETH Z\"{u}rich;Ambrosio, Luigi,2005

2. C. Andrieu, N. De Freitas, A. Doucet, and M. I. Jordan, An introduction to MCMC for machine learning, Mach. Learn. 50 (2003), 5–43.

3. M. Arjovsky, S. Chintala, and L. Bottou, Wasserstein generative adversarial networks, in International conference on machine learning, PMLR, 2017, pp. 214–223.

4. R. Baptista, Y. Marzouk, O. Zahm, On the representation and learning of monotone triangular transport maps, Found. Comput. Math. (2023), 1–46, DOI:10.1007/s10208-023-09630-x.

5. A computational fluid mechanics solution to the Monge-Kantorovich mass transfer problem;Benamou, Jean-David;Numer. Math.,2000

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3