Lipschitz-regularized gradient flows and generative particle algorithms for high-dimensional scarce data

DOI10.1137/23M1587841MaRDI QIDQ6664473zbMATH OpenFDO

Authors Hyemin Gu, P. Birmpa, Yannis Pantazis, Luc Rey-Bellet, Markos Katsoulakis

Publication date 16 January 2025

Published in SIAM Journal on Mathematics of Data Science (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1137/23M1587841

zbMATH Keywords

data integrationoptimal transportgradient flowinformation theoryparticle algorithmsgenerative modeling

Mathematics Subject Classification ID

Statistical aspects of information-theoretic topics (62B10) Optimal transportation (49Q22) Artificial neural networks and deep learning (68T07) Computational learning theory (68Q32) Measures of information, entropy (94A17) Fokker-Planck equations (35Q84) Stochastic particle methods (65C35) Flows in porous media; filtration; seepage (76S05)

Recommendations

• Projected Wasserstein Gradient Descent for High-Dimensional Bayesian Inference
• Minimax rate of distribution estimation on unknown submanifolds under adversarial losses
• Neural parametric Fokker-Planck equation
• An algorithm for learning representations of models with scarce data

Cites work

• scientific article; zbMATH DE number 2152346 (Why is no real title available?)
• scientific article; zbMATH DE number 5180707 (Why is no real title available?)
• Barenblatt solutions and asymptotic behaviour for a nonlinear fractional heat equation of porous medium type
• Estimation of non-normalized statistical models by score matching
• Exponential convergence of Langevin distributions and their discrete approximations
• Fokker-Planck particle systems for Bayesian inference: computational approaches
• Formulation and properties of a divergence used to compare probability measures without absolute continuity
• Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality
• Lecture notes on the DiPerna-Lions theory in abstract measure spaces
• Nonasymptotic convergence analysis for the unadjusted Langevin algorithm
• Numerical study of a particle method for gradient flows
• Optimizing Variational Representations of Divergences and Accelerating Their Statistical Estimation
• Ordinary differential equations, transport theory and Sobolev spaces
• Pattern recognition and machine learning.
• Principal component analysis: a review and recent developments
• Scaling Limit of the Stein Variational Gradient Descent: The Mean Field Regime
• THE GEOMETRY OF DISSIPATIVE EVOLUTION EQUATIONS: THE POROUS MEDIUM EQUATION
• The Variational Formulation of the Fokker--Planck Equation
• The elements of statistical learning. Data mining, inference, and prediction
• (f,)-divergences: interpolating between f-divergences and integral probability metrics
• \(L^q\)-functional inequalities and weighted porous media equations

Cited in

(3)

• Robust generative learning with Lipschitz-regularized -divergences allows minimal assumptions on target distributions
• An algorithm for learning representations of models with scarce data
• Nonlinear denoising score matching for enhanced learning of structured distributions

This page was built for publication: Lipschitz-regularized gradient flows and generative particle algorithms for high-dimensional scarce data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6664473)