Multi-stage splitting integrators for sampling with modified Hamiltonian Monte Carlo methods

Abstract

Modified Hamiltonian Monte Carlo (MHMC) methods combine the ideas behind two popular sampling approaches: Hamiltonian Monte Carlo (HMC) and importance sampling. As in the HMC case, the bulk of the computational cost of MHMC algorithms lies in the numerical integration of a Hamiltonian system of differential equations. We suggest novel integrators designed to enhance accuracy and sampling performance of MHMC methods. The novel integrators belong to families of splitting algorithms and are therefore easily implemented. We identify optimal integrators within the families by minimizing the energy error or the average energy error. We derive and discuss in detail the modified Hamiltonians of the new integrators, as the evaluation of those Hamiltonians is key to the efficiency of the overall algorithms. Numerical experiments show that the use of the new integrators may improve very significantly the sampling performance of MHMC methods, in both statistical and molecular dynamics problems.

arXiv comment31 pages, 9 figures. arXiv admin note: text overlap with arXiv:1706.04032

References in corpus (11)

• On the construction of high-order force gradient algorithms for integration of motion in classical and quantum systems
• Geometric integrators and the Hamiltonian Monte Carlo method
• Numerical integrators for the Hybrid Monte Carlo method
• Emulation of Higher-Order Tensors in Manifold Monte Carlo Methods for Bayesian Inverse Problems
• Modified Hamiltonian Monte Carlo for Bayesian inference
• Adaptive multi-stage integrators for optimal energy conservation in molecular simulations
• Magnetic Hamiltonian Monte Carlo
• Palindromic 3-stage splitting integrators, a roadmap
• Geometrically Tempered Hamiltonian Monte Carlo
• On an asymptotic method for computing the modified energy for symplectic methods
• Enhancing sampling in atomistic simulations of solid state materials for batteries: a focus on olivine NaFePO4

Cited by in corpus (7)

• Modified Hamiltonian Monte Carlo for Bayesian inference
• Geometric Methods for Sampling, Optimisation, Inference and Adaptive Agents
• Sampling algorithms in statistical physics: a guide for statistics and machine learning
• Adaptive multi-stage integration schemes for Hamiltonian Monte Carlo
• HMC: avoiding rejections by not using leapfrog and some results on the acceptance rate
• Is the NUTS algorithm correct?
• Symmetrically processed splitting integrators for enhanced Hamiltonian Monte Carlo sampling