Control of probability flow in Markov chain Monte Carlo -- Nonreversibility and lifting
arXiv:1207.0258 · doi:10.1063/5.0233858
Abstract
The Markov chain Monte Carlo (MCMC) method is widely used in various fields as a powerful numerical integration technique for systems with many degrees of freedom. In MCMC methods, probabilistic state transitions can be considered as a random walk in state space, and random walks allow for sampling from complex distributions. However, paradoxically, it is necessary to carefully suppress the randomness of the random walk to improve computational efficiency. By breaking detailed balance, we can create a probability flow in the state space and perform more efficient sampling along this flow. Motivated by this idea, practical and efficient nonreversible MCMC methods have been developed over the past ten years. In particular, the lifting technique, which introduces probability flows in an extended state space, has been applied to various systems and has proven more efficient than conventional reversible updates. We review and discuss several practical approaches to implementing nonreversible MCMC methods, including the shift method in the cumulative distribution and the directed-worm algorithm.
arXiv comment10 pages, 9 figures
References in corpus (35)
- Quantum Monte Carlo with Directed Loops
- Worm Algorithm and Diagrammatic Monte Carlo: A New Approach to Continuous-Space Path Integral Monte Carlo Simulations
- Exact, Complete, and Universal Continuous-Time Worldline Monte Carlo Approach to the Statistics of Discrete Quantum Systems
- Worm algorithms for classical statistical models
- Hard-disk equation of state: First-order liquid-hexatic transition in two dimensions with three simulation methods
- Event-chain Monte Carlo algorithms for hard-sphere systems
- Markov Chain Monte Carlo Method without Detailed Balance
- Generalized event-chain Monte Carlo: Constructing rejection-free global-balance algorithms from infinitesimal steps
- Irreversible Monte Carlo Algorithms for Efficient Sampling
- Variance Reduction using Nonreversible Langevin Samplers
- Accelerating diffusions
- Dynamic scaling at classical phase transitions approached through non-equilibrium quenching
- Non-reversible Metropolis-Hastings
- Non-Reversible Parallel Tempering: a Scalable Highly Parallel MCMC Scheme
- Chiral Limit of Strongly Coupled Lattice Gauge Theories
- Violation of detailed balance accelerates relaxation
- Event-chain Monte Carlo: foundations, applications, and prospects
- Hard-sphere melting and crystallization with event-chain Monte Carlo
- Event-chain Monte Carlo for classical continuous spin models
- Event-chain Monte Carlo algorithms for three- and many-particle interactions
- Event-chain algorithm for the Heisenberg model: Evidence for dynamic scaling
- Dynamics of the One-Dimensional Ising Model without Detailed Balance Condition
- Cell-veto Monte Carlo algorithm for long-range systems
- Worm Monte Carlo study of the honeycomb-lattice loop model
- Role of current fluctuations in nonreversible samplers
- Efficient Irreversible Monte Carlo samplers
- A Worm Algorithm for Two-Dimensional Spin Glasses
- Geometric Allocation Approaches in Markov Chain Monte Carlo
- Irreversible simulated tempering
- Multithreaded event-chain Monte Carlo with local times
- Regeneration-enriched Markov processes with application to Monte Carlo
- Lifted Worm Algorithm for the Ising Model
- Geometric allocation approach to accelerating directed worm algorithm
- Lifted directed-worm algorithm
- Reducing rejection exponentially improves Markov chain Monte Carlo sampling