Hypocoercivity in metastable settings and kinetic simulated annealing
Pierre Monmarch\'e

TL;DR
This paper proves the convergence over time of kinetic Langevin annealing with a logarithmic cooling schedule by integrating classical simulated annealing analysis with hypocoercive methods, advancing understanding of metastable systems.
Contribution
It introduces a novel combination of classical and hypocoercive techniques to analyze convergence in metastable kinetic annealing processes.
Findings
Proves convergence of kinetic Langevin annealing with logarithmic cooling.
Integrates classical annealing analysis with hypocoercive methods.
Provides theoretical foundation for metastable system behavior.
Abstract
Combining classical arguments for the analysis of the simulated annealing algorithm with the more recent hypocoercive method of distorted entropy, we prove the convergence for large time of the kinetic Langevin annealing with logarithmic cooling schedule.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsStatistical Mechanics and Entropy · Markov Chains and Monte Carlo Methods · Advanced Thermodynamics and Statistical Mechanics
