Sampling and statistical physics via symmetry

Abstract

We formulate both Markov chain Monte Carlo (MCMC) sampling algorithms and basic statistical physics in terms of elementary symmetries. This perspective on sampling yields derivations of well-known MCMC algorithms and a new parallel algorithm that appears to converge more quickly than current state of the art methods. The symmetry perspective also yields a parsimonious framework for statistical physics and a practical approach to constructing meaningful notions of effective temperature and energy directly from time series data. We apply these latter ideas to Anosov systems.

arXiv commentProceedings of Les Houches 2020 school on Joint Structures and Common Foundations of Statistical Physics, Information Geometry and Inference for Learning

References in corpus (35)

• MCMC using Hamiltonian dynamics
• Parallel Tempering: Theory, Applications, and New Perspectives
• Dynamical ensembles in stationary states
• A Dynamical Approach to Temperature
• The effective temperature
• Negative Absolute Temperature for Motional Degrees of Freedom
• Von Neumann Algebra Automorphisms and Time-Thermodynamics Relation in General Covariant Quantum Theories
• Temperature in and out of equilibrium: a review of concepts, tools and attempts
• Inconsistent thermostatistics and negative absolute temperatures
• Markov Chain Monte Carlo Method without Detailed Balance
• Efficient computation of the Zassenhaus formula
• A Characterization of Entropy in Terms of Information Loss
• Gibbs, Boltzmann, and negative temperatures
• Thermodynamics of the glassy state: effective temperature as an additional system parameter
• Diamonds's Temperature: Unruh effect for bounded trajectories and thermal time hypothesis
• Thermal time and the Tolman-Ehrenfest effect: temperature as the "speed of time"
• Chaotic Hypothesis, Fluctuation Theorem and singularities
• A Review of Multiple Try MCMC algorithms for Signal Processing
• Proceedings of the Twenty-Ninth Conference on Uncertainty in Artificial Intelligence (2013)
• A Geometric, Dynamical Approach to Thermodynamics
• Non-reversible Metropolis-Hastings
• The Sherrington-Kirkpatrick model: an overview
• Optimal Monte Carlo Updating
• Heat and Fluctuations from Order to Chaos
• Analysis and Optimization of Population Annealing
• Nonequilibrium, thermostats and thermodynamic limit
• Optimization of Markov process violates detailed balance condition
• De Sitter Thermodynamics from Diamonds's Temperature
• Open Sets of Exponentially Mixing Anosov Flows
• Smooth mixing Anosov flows in dimension three are exponential mixing
• On Thermostats: Isokinetic or Hamiltonian? finite or infinite?
• Thermodynamic limit for isokinetic thermostats
• Analyticity of the SRB measure for a class of simple Anosov flows
• On the stochastic Lie algebra
• Descriptive Thermodynamics