Quantization of probability densities: A gradient flow approach

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper introduces a gradient flow in infinite dimension, whose long-time dynamics is expected to be an approximation of the quantization problem for probability densities, in the sense of Graf and Luschgy (Lecture Notes in Mathematics, vol 1730. Springer, Berlin, 2000). Quantization of probability distributions is a problem which one encounters in a great variety of contexts, such as signal processing, pattern or speech recognition, economics.. The present work describes a dynamical approach of the optimal quantization problem in space dimensions one and two, involving (systems of) parabolic equations. This is an account of recent work in collaboration with Caglioti et al. (Math Models Methods Appl Sci 25:1845–1885, 2015 and arXiv:1607.01198 (math.AP), to appear in Ann. Inst. H. Poincaré, Anal. Non Lin. https://doi.org/10.1016/j.anihpc.2017.12.003).

Original languageEnglish
Title of host publicationFrom Particle Systems to Partial Differential Equations - PSPDE V 2016
EditorsPatrícia Gonçalves, Ana Jacinta Soares
PublisherSpringer New York LLC
Pages33-52
Number of pages20
ISBN (Print)9783319996882
DOIs
Publication statusPublished - 1 Jan 2018
Event5th international conference on Particle Systems and Partial Differential Equations, PS-PDEs V 2016 - Braga, Portugal
Duration: 28 Nov 201630 Nov 2016

Publication series

NameSpringer Proceedings in Mathematics and Statistics
Volume258
ISSN (Print)2194-1009
ISSN (Electronic)2194-1017

Conference

Conference5th international conference on Particle Systems and Partial Differential Equations, PS-PDEs V 2016
Country/TerritoryPortugal
CityBraga
Period28/11/1630/11/16

Keywords

  • Gradient flow
  • Parabolic equations
  • Quantization of probability densities
  • Wasserstein distance

Fingerprint

Dive into the research topics of 'Quantization of probability densities: A gradient flow approach'. Together they form a unique fingerprint.

Cite this