Adaptive Thermostats for Noisy Gradient Systems

Abstract

We study numerical methods for sampling probability measures in high dimension where the underlying model is only approximately identified with a gradient system. Extended stochastic dynamical methods are discussed which have application to multiscale models, nonequilibrium molecular dynamics, and Bayesian sampling techniques arising in emerging machine learning applications. In addition to providing a more comprehensive discussion of the foundations of these methods, we propose a new numerical method for the adaptive Langevin/stochastic gradient Nosé--Hoover thermostat that achieves a dramatic improvement in numerical efficiency over the most popular stochastic gradient methods reported in the literature. We also demonstrate that the newly established method inherits a superconvergence property (fourth order convergence to the invariant measure for configurational quantities) recently demonstrated in the setting of Langevin dynamics. Our findings are verified by numerical experiments.

References in corpus (3)

• Comparison of Dissipative Particle Dynamics and Langevin thermostats for out-of-equilibrium simulations of polymeric systems
• On the numerical treatment of dissipative particle dynamics and related systems
• The Generalized Langevin Equation: An efficient approach to non-equilibrium molecular dynamics of open systems

Cited by in corpus (13)

• Structure-preserving integrators for dissipative systems based on reversible-irreversible splitting
• Pairwise adaptive thermostats for improved accuracy and stability in dissipative particle dynamics
• Assessing numerical methods for molecular and particle simulation
• Irreversible Samplers from Jump and Continuous Markov Processes
• Accurate and robust splitting methods for the generalized Langevin equation with a positive Prony series memory kernel
• Simplest random walk for approximating Robin boundary value problems and ergodic limits of reflected diffusions
• Bayesian Learning of Parameterised Quantum Circuits
• Time correlation functions of equilibrium and nonequilibrium Langevin dynamics: Derivations and numerics using random numbers
• Generalized Bayesian Likelihood-Free Inference
• Dynamic principle for ensemble control tools
• Accurate and efficient splitting methods for dissipative particle dynamics
• Stochastic Norton dynamics: An alternative approach for the computation of transport coefficients in dissipative particle dynamics
• Sharp spectral gap of adaptive Langevin dynamics