Mirror descent algorithms for minimizing interacting free energy

DOI10.1007/S10915-020-01303-ZMaRDI QIDQ2204545zbMATH OpenOpenAlexFDO

Authors Lexing Ying

Publication date 15 October 2020

Published in Journal of Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2004.04555

zbMATH Keywords

Hellinger divergenceinteracting free energyKullback-Leibler divergencemirror descent algorithms

Mathematics Subject Classification ID

Numerical methods based on nonlinear programming (49M37) Numerical optimization and variational techniques (65K10)

Abstract: This note considers the problem of minimizing interacting free energy. Motivated by the mirror descent algorithm, for a given interacting free energy, we propose a descent dynamics with a novel metric that takes into consideration the reference measure and the interacting term. This metric naturally suggests a monotone reparameterization of the probability measure. By discretizing the reparameterized descent dynamics with the explicit Euler method, we arrive at a new mirror-descent-type algorithm for minimizing interacting free energy. Numerical results are included to demonstrate the efficiency of the proposed algorithms.

Recommendations

• Hessian informed mirror descent
• On stochastic mirror descent with interacting particles: convergence properties and variance reduction
• Mirror descent and convex optimization problems with non-smooth inequality constraints
• Mirror descent and nonlinear projected subgradient methods for convex optimization.
• The information geometry of mirror descent

Cites work

• A blob method for diffusion
• A finite volume scheme for nonlinear degenerate parabolic equations
• A mean field view of the landscape of two-layer neural networks
• Convex optimization: algorithms and complexity
• scientific article; zbMATH DE number 3790208 (Why is no real title available?)
• scientific article; zbMATH DE number 3562346 (Why is no real title available?)
• Information geometry and its applications
• Kinetic equilibration rates for granular media and related equations: entropy dissipation and mass transportation estimates
• Mathematics of granular materials
• Mean field analysis of neural networks: a central limit theorem
• Mirror descent and nonlinear projected subgradient methods for convex optimization.
• Natural gradient via optimal transport
• Positivity-preserving and asymptotic preserving method for 2D Keller-Segal equations
• THE GEOMETRY OF DISSIPATIVE EVOLUTION EQUATIONS: THE POROUS MEDIUM EQUATION
• The multiplicative weights update method: a meta-algorithm and applications
• The Variational Formulation of the Fokker--Planck Equation

Cited in

(5)

• Natural gradient for combined loss using wavelets
• On stochastic mirror descent with interacting particles: convergence properties and variance reduction
• Hessian informed mirror descent
• Particle dual averaging: optimization of mean field neural network with global convergence rate analysis*
• Approximate Newton Policy Gradient Algorithms

This page was built for publication: Mirror descent algorithms for minimizing interacting free energy

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2204545)