Uniform decomposition of probability measures: quantization, clustering and rate of convergence

DOI10.1017/JPR.2018.69zbMATH Open1405.60025arXiv1801.02871OpenAlexW2909936794WikidataQ128576312 ScholiaQ128576312MaRDI QIDQ4611267FDOQ4611267

Authors: Julien Chevallier 

Publication date: 17 January 2019

Published in: Journal of Applied Probability (Search for Journal in Brave)

Abstract: The study of finite approximations of probability measures has a long history. In (Xu and Berger, 2017), the authors focus on constrained finite approximations and, in particular, uniform ones in dimension $d=1$. The present paper gives an elementary construction of a uniform decomposition of probability measures in dimension $dgeq1$. This decomposition is then used to give upper-bounds on the rate of convergence of the optimal uniform approximation error. These bounds appear to be the generalization of the ones obtained in (Xu and Berger, 2017) and to be sharp for generic probability measures.

Full work available at URL: https://arxiv.org/abs/1801.02871

Recommendations

• Constructive quantization: approximation by empirical measures
• scientific article; zbMATH DE number 7688072
• Quantization for probability measures in the Prokhorov metric
• Optimal approximation and quantisation
• Optimal quantization for piecewise uniform distributions

zbMATH Keywords

clusteringrate of convergenceWasserstein distancequantizationuniform approximation

Mathematics Subject Classification ID

Convergence of probability measures (60B10) Inequalities; stochastic orderings (60E15) Approximations to statistical distributions (nonasymptotic) (62E17) Limit theorems in probability theory (60F99)

Cites Work

• Foundations of quantization for probability distributions
• Approximation by finitely supported measures
• On the rate of convergence in Wasserstein distance of the empirical measure
• On the mean speed of convergence of empirical and occupation measures in Wasserstein distance
• Asymptotic quantization error of continuous signals and the quantization dimension
• High-dimensional integration: The quasi-Monte Carlo way
• Quantitative concentration inequalities for empirical measures on non-compact spaces
• Optimal Delaunay and Voronoi quantization schemes for pricing American style options
• Random bit quadrature and approximation of distributions on Hilbert spaces
• Introduction to vector quantization and its applications for numerics
• Mean field limits for nonlinear spatially extended Hawkes processes with exponential memory kernels
• Support points

Cited In (11)

• Randomized Wasserstein barycenter computation: resampling with statistical guarantees
• Curve based approximation of measures on manifolds by discrepancy minimization
• Asymptotics of one-dimensional Lévy approximations
• Best finite constrained approximations of one-dimensional probabilities
• From optimal transport to discrepancy
• Random bit quadrature and approximation of distributions on Hilbert spaces
• Designing universal causal deep learning models: The geometric (Hyper)transformer
• Approximation rate in Wasserstein distance of probability measures on the real line by deterministic empirical measures
• Universal regular conditional distributions via probabilistic transformers
• Random bit multilevel algorithms for stochastic differential equations
• Vlasov equations on digraph measures

This page was built for publication: Uniform decomposition of probability measures: quantization, clustering and rate of convergence

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4611267)