Notes on the Hybrid Monte Carlo Method

Jeremy C. Palmer,Amir Haji-Akbari,Rakesh S. Singh,Fausto Martelli,Roberto Car,Athanassios Z. Panagiotopoulos,Pablo G. Debenedetti

DOI: https://doi.org/10.48550/arXiv.1712.08278

2017-12-22

Abstract:We discuss the detailed balance condition for hybrid Monte Carlo method

Statistical Mechanics

What problem does this paper attempt to address?

An Improved Acceptance Procedure for the Hybrid Monte Carlo Algorithm

R. M. Neal

DOI: https://doi.org/10.48550/arXiv.hep-lat/9208011

1992-08-21

Abstract:The probability of accepting a candidate move in the hybrid Monte Carlo algorithm can be increased by considering a transition to be between windows of several states at the beginning and end of the trajectory, with a state within the selected window being chosen according to the Boltzmann probabilities. The detailed balance condition used to justify the algorithm still holds with this procedure, provided the start state is randomly positioned within its window. The new procedure is shown empirically to significantly improve performance for a test system of uncoupled oscillators.

High Energy Physics - Lattice
A Multiscale Hybrid Method

Gabriel R. Barrenechea,Antonio Tadeu A. Gomes,Diego Paredes

DOI: https://doi.org/10.1137/22m1542556

2024-05-13

Theory of Probability and Its Applications

Abstract:SIAM Journal on Scientific Computing, Volume 46, Issue 3, Page A1628-A1657, June 2024. In this work we propose, analyze, and test a new multiscale finite element method called Multiscale Hybrid (MH) method. The method is built as a close relative to the Multiscale Hybrid Mixed (MHM) method, but with the fundamental difference that a novel definition of the Lagrange multiplier is introduced. The practical implication of this is that both the local problems to compute the basis functions, as well as the global problem, are elliptic, as opposed to the MHM method (and also other previous methods) where a mixed global problem is solved and constrained local problems are solved to compute the local basis functions. The error analysis of the method is based on a hybrid formulation, and a static condensation process is done at the discrete level, so the final global system only involves the Lagrange multipliers. We tested the performance of the method by means of numerical experiments for problems with multiscale coefficients, and we carried out comparisons with the MHM method in terms of performance, accuracy, and memory requirements.

statistics & probability
"Parallel-Tempering"-Assisted Hybrid Monte Carlo Algorithm For Bayesian Inference In Dynamical Systems

Shengjie Sun,Yuan Shen

DOI: https://doi.org/10.1007/978-3-030-29933-0_30

2020-01-01

Abstract:The aim of this work is to tackle the problem of sampling from multi-modal distributions when Hybrid Monte Carlo (HMC) algorithm is employed for performing Bayesian inference in dynamical systems. Hybrid Monte Carlo is a powerful Markov Chain Monte Carlo (MCMC) algorithm but it still suffers from the "multiple peaks" problem. Due to non-trivial structure in the space of (a class of) dynamical systems, posterior distribution of its model parameters could exhibit complicated structures such as multiple ridges. We examined a MCMC algorithm combining HMC with so-called Parallel Tempering (PT) - a well-known strategy for tackling the problem highlighted above. The new algorithm is referred to as PT-HMC. Our numerical experiment demonstrated that when compared to the ground truth, the posterior distributions derived from PT-HMC samples is more accurate than those from HMC.
Issues in designing hybrid algorithms

Jeong Lee,Kerrie Mengersen,Christian Robert,Ross McVinish

DOI: https://doi.org/10.48550/arXiv.1111.2609

2011-11-11

Abstract:In the Bayesian community, an ongoing imperative is to develop efficient algorithms. An appealing approach is to form a hybrid algorithm by combining ideas from competing existing techniques. This paper addresses issues in designing hybrid methods by considering selected case studies: the delayed rejection algorithm, the pinball sampler, the Metropolis adjusted Langevin algorithm, and the population Monte Carlo algorithm. We observe that even if each component of a hybrid algorithm has individual strengths, they may not contribute equally or even positively when they are combined. Moreover, even if the statistical efficiency is improved, from a practical perspective there are technical issues to be considered such as applicability and computational workload. In order to optimize performance of the algorithm in real time, these issues should be taken into account.

Methodology
Hybrid Monte Carlo-Methods in Credit Risk Management

Lucia Del Chicca,Gerhard Larcher

DOI: https://doi.org/10.48550/arXiv.1405.1831

2014-05-08

Numerical Analysis

Abstract:In this paper we analyze and compare the use of Monte Carlo, Quasi-Monte Carlo and hybrid Monte Carlo-methods in the credit risk management system Credit Metrics by J.P.Morgan. We show that hybrid sequences used for simulations, in a suitable way, in many relevant situations, perform better than pure Monte Carlo and pure Quasi-Monte Carlo methods, and they essentially never perform worse than these methods.
An irreversible Markov-chain Monte Carlo method with skew detailed balance conditions

K Hukushima,Y Sakai

DOI: https://doi.org/10.1088/1742-6596/473/1/012012

2013-12-16

Journal of Physics: Conference Series

Abstract:An irreversible Markov-chain Monte Carlo (MCMC) method based on a skew detailed balance condition is discussed. Some recent theoretical works concerned with the irreversible MCMC method are reviewed and the irreversible Metropolis-Hastings algorithm for the method is described. We apply the method to ferromagnetic Ising models in two and three dimensions. Relaxation dynamics of the order parameter and the dynamical exponent are studied in comparison to those with the conventional reversible MCMC method with the detailed balance condition. We also examine how the efficiency of exchange Monte Carlo method is affected by the combined use of the irreversible MCMC method.
A Hybrid Monte Carlo, Discontinuous Galerkin method for linear kinetic transport equations

Johannes Krotz,Cory D. Hauck,Ryan G. McClarren

2023-12-07

Abstract:We present a hybrid method for time-dependent particle transport problems that combines Monte Carlo (MC) estimation with deterministic solutions based on discrete ordinates. For spatial discretizations, the MC algorithm computes a piecewise constant solution and the discrete ordinates uses bilinear discontinuous finite elements. From the hybridization of the problem, the resulting problem solved by Monte Carlo is scattering free, resulting in a simple, efficient solution procedure. Between time steps, we use a projection approach to ``relabel'' collided particles as uncollided particles. From a series of standard 2-D Cartesian test problems we observe that our hybrid method has improved accuracy and reduction in computational complexity of approximately an order of magnitude relative to standard discrete ordinates solutions.

Numerical Analysis
Analysis of Hybrid MC/Deterministic Methods for Transport Problems Based on Low-Order Equations Discretized by Finite Volume Schemes

Vincent N. Novellino,Dmitriy Y. Anistratov

2024-03-12

Abstract:This paper presents hybrid numerical techniques for solving the Boltzmann transport equation formulated by means of low-order equations for angular moments of the angular flux. The moment equations are derived by the projection operator approach. The projected equations are closed exactly using a high-order transport solution. The low-order equations of the hybrid methods are approximated with a finite volume scheme of the second-order accuracy. Functionals defining the closures in the discretized low-order equations are calculated by Monte Carlo techniques. In this study, we analyze effects of statistical noise and discretization error on the accuracy of the hybrid transport solution.

Numerical Analysis
A hybrid method of improving source convergence implemented in RMC

Ding She,Zeguang Li,Qi Xu,Kan Wang,Ganglin Yu

2011-01-01

Abstract:Based on the fission matrix method and Wielandts method, a new hybrid method is proposed to improve the fission source convergence in MC criticality calculation. The hybrid method is implemented into the MC code RMC. A loosely coupled benchmark and a large scale problem are calculated, which demonstrate that the new hybrid method is effective for improving source convergence. Further research would be focused on optimizing the parameters of the hybrid method, and convergence diagnostics.
Hybridization and postprocessing in finite element exterior calculus

Gerard Awanou,Maurice Fabien,Johnny Guzmán,Ari Stern

DOI: https://doi.org/10.48550/arXiv.2008.00149

2021-10-07

Abstract:We hybridize the methods of finite element exterior calculus for the Hodge-Laplace problem on differential $k$-forms in $\mathbb{R}^n$. In the cases $k = 0$ and $k = n$, we recover well-known primal and mixed hybrid methods for the scalar Poisson equation, while for $0 < k < n$, we obtain new hybrid finite element methods, including methods for the vector Poisson equation in $n = 2$ and $n = 3$ dimensions. We also generalize Stenberg postprocessing from $k = n$ to arbitrary $k$, proving new superconvergence estimates. Finally, we discuss how this hybridization framework may be extended to include nonconforming and hybridizable discontinuous Galerkin methods.

Numerical Analysis
A Hybrid Stochastic Gradient Hamiltonian Monte Carlo Method

Chao Zhang,Zhijian Li,Zebang Shen,Jiahao Xie,Hui Qian

DOI: https://doi.org/10.1609/aaai.v35i12.17295

2021-01-01

Proceedings of the AAAI Conference on Artificial Intelligence

Abstract:Recent theoretical analyses reveal that existing Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) methods need large mini-batches of samples (exponentially dependent on the dimension) to reduce the mean square error of gradient estimates and ensure non-asymptotic convergence guarantees when the target distribution has a nonconvex potential function. In this paper, we propose a novel SG-MCMC algorithm, called Hybrid Stochastic Gradient Hamiltonian Monte Carlo (HSG-HMC) method, which needs merely one sample per iteration and possesses a simple structure with only one hyperparameter. Such improvement leverages a hybrid stochastic gradient estimator that exploits historical stochastic gradient information to control the mean square error. Theoretical analyses show that our method obtains the best-known overall sample complexity to achieve epsilon-accuracy in terms of the 2-Wasserstein distance for sampling from distributions with nonconvex potential functions. Empirical studies on both simulated and real-world datasets demonstrate the advantage of our method.
Non-Hermitian Polynomial Hybrid Monte Carlo

Oliver Witzel

DOI: https://doi.org/10.48550/arXiv.0809.1010

2008-09-05

Abstract:We report on a new variant of the hybrid Monte Carlo algorithm employing a polynomial approximation of the inverse of the non-Hermitian Dirac-Wilson operator. Our approximation relies on simple and stable recurrence relations of complex Chebyshev polynomials. First performance figures are presented.

High Energy Physics - Lattice
Overcoming Ergodicity Problems of the Hybrid Monte Carlo Method using Radial Updates

Finn Temmen,Evan Berkowitz,Anthony Kennedy,Thomas Luu,Johann Ostmeyer,Xinhao Yu

2024-10-25

Abstract:Despite its many advantages, the sensible application of the Hybrid Monte Carlo (HMC) method is often hindered by the presence of large - or even infinite - potential barriers. These potential barriers partition the configuration space into distinct sectors, which leads to ergodicity violations and biased measurements of observables. In this work, we address this problem by augmenting the HMC method with a multiplicative Metropolis-Hastings update in a so-called "radial direction" of the fields, which enables jumps over the aforementioned potential barriers at comparably low computational cost. The effectiveness of this approach is demonstrated for the Hubbard model, formulated in a non-compact space by means of a continuous Hubbard-Stratonovich transformation. Our numerical results show that the radial updates successfully resolve the ergodicity violation, while simultaneously reducing autocorrelations.

Strongly Correlated Electrons
Optimized Monte Carlo Methods

Enzo Marinari

DOI: https://doi.org/10.48550/arXiv.cond-mat/9612010

1996-12-01

Abstract:I discuss optimized data analysis and Monte Carlo methods. Reweighting methods are discussed through examples, like Lee-Yang zeroes in the Ising model and the absence of deconfinement in QCD. I discuss reweighted data analysis and multi-hystogramming. I introduce Simulated Tempering, and as an example its application to the Random Field Ising Model. I illustrate Parallel Tempering, and discuss some technical crucial details like thermalization and volume scaling. I give a general perspective by discussing Umbrella Methods and the Multicanonical approach.

Disordered Systems and Neural Networks,Statistical Mechanics,High Energy Physics - Lattice
An Efficient Method for Treating Conditional Monte Carlo Simulation

HP FANG

DOI: https://doi.org/10.1016/0010-4655(94)90044-2

IF: 4.717

1994-01-01

Computer Physics Communications

Abstract:One of the most commonly used methods to treat the conditional Monte Carlo is to transfer the conditional probability distributions to unconditional probability distributions with properly compensated weighting factors. Generally, the weighting factors are “awkward” functions in high-dimensional systems. In this paper, a technique is proposed to improve this method by getting rid of the “awkward” parts of the weighting factors.
Hybrid Auxiliary Field Quantum Monte Carlo for Molecular Systems

Yixiao Chen,Linfeng Zhang,Weinan E,Roberto Car

DOI: https://doi.org/10.1021/acs.jctc.3c00038

2023-04-19

Journal of Chemical Theory and Computation

Abstract:We propose a quantum Monte Carlo approach to solve the many-body Schrödinger equation for the electronic ground state. The method combines optimization from variational Monte Carlo and propagation from auxiliary field quantum Monte Carlo in a way that significantly alleviates the sign problem. In application to molecular systems, we obtain highly accurate results for configurations dominated by either dynamic or static electronic correlation.

chemistry, physical,physics, atomic, molecular & chemical
Accelerating the Hybrid Monte Carlo algorithm with ILU preconditioning

Mike Peardon

DOI: https://doi.org/10.48550/arXiv.hep-lat/0011080

2000-12-13

Abstract:The pseudofermion action of the Hybrid Monte Carlo (HMC) algorithm for dynamical fermions is modified to directly incorporate Incomplete LU (ILU) factorisation. This reduces the stochastic noise and allows a larger molecular dynamics step-size to be taken, cutting the computational cost. Numerical tests using the two-flavour Schwinger model are presented, where a two-step ILU preconditioning of the even-odd fermion matrix allows the step-size to be increased by a factor of two over the standard even-odd formulation.

High Energy Physics - Lattice
The Geometry of Hamiltonian Monte Carlo

Michael Betancourt,Leo C. Stein

DOI: https://doi.org/10.48550/arXiv.1112.4118

2011-12-18

Methodology

Abstract:With its systematic exploration of probability distributions, Hamiltonian Monte Carlo is a potent Markov Chain Monte Carlo technique; it is an approach, however, ultimately contingent on the choice of a suitable Hamiltonian function. By examining both the symplectic geometry underlying Hamiltonian dynamics and the requirements of Markov Chain Monte Carlo, we construct the general form of admissible Hamiltonians and propose a particular choice with potential application in Bayesian inference.
Fluid Solver Independent Hybrid Methods for Multiscale Kinetic equations

Giacomo Dimarco,Lorenzo Pareschi

DOI: https://doi.org/10.48550/arXiv.0908.0416

2009-08-04

Abstract:In some recent works [G. Dimarco, L. Pareschi, Hybrid multiscale methods I. Hyperbolic Relaxation Problems, Comm. Math. Sci., 1, (2006), pp. 155-177], [G. Dimarco, L. Pareschi, Hybrid multiscale methods II. Kinetic equations, SIAM Multiscale Modeling and Simulation Vol 6., No 4,pp. 1169-1197, (2008)] we developed a general framework for the construction of hybrid algorithms which are able to face efficiently the multiscale nature of some hyperbolic and kinetic problems. Here, at variance with respect to the previous methods, we construct a method form-fitting to any type of finite volume or finite difference scheme for the reduced equilibrium system. Thanks to the coupling of Monte Carlo techniques for the solution of the kinetic equations with macroscopic methods for the limiting fluid equations, we show how it is possible to solve multiscale fluid dynamic phenomena faster with respect to traditional deterministic/stochastic methods for the full kinetic equations. In addition, due to the hybrid nature of the schemes, the numerical solution is affected by less fluctuations when compared to standard Monte Carlo schemes. Applications to the Boltzmann-BGK equation are presented to show the performance of the new methods in comparison with classical approaches used in the simulation of kinetic equations.

Numerical Analysis,Mathematical Physics,Computational Physics,Fluid Dynamics
Hybrid Monte Carlo estimators for multilayer transport problems

Shuang Zhao,Jerome Spanier

DOI: https://doi.org/10.1016/j.jcp.2021.110117

IF: 4.645

2021-04-01

Journal of Computational Physics

Abstract:This paper introduces a new family of hybrid estimators aimed at controlling the efficiency of Monte Carlo computations in particle transport problems. In this context, efficiency is usually measured by the figure of merit (FOM) given by the inverse product of the estimator variance <math>Var[ξ]</math> and the run time T: <math>FOM:=(Var[ξ]T)−1</math>. Previously, we developed a new family of transport-constrained unbiased radiance estimators (T-CURE) that generalize the conventional collision and track length estimators <a class="workspace-trigger" href="#br0010">[1]</a> and provide 1–2 orders of magnitude additional variance reduction. However, these gains in variance reduction are partly offset by increases in overhead time <a class="workspace-trigger" href="#br0020">[2]</a>, lowering their computational efficiency. Here we show that combining T-CURE estimation with conventional terminal estimation within each individual biography can moderate the efficiency of the resulting "hybrid" estimator without introducing bias in the computation. This is achieved by treating only the refractive interface crossings with the extended next event estimator, and all others by standard terminal estimators. This is because when there are index-mismatched interfaces between the collision location and the detector, the T-CURE computation rapidly becomes intractable due to the large number of refractions and reflections that can arise. We illustrate the gains in efficiency by comparing our hybrid strategy with more conventional estimation methods in a series of multi-layer numerical examples.

physics, mathematical,computer science, interdisciplinary applications

Notes on the Hybrid Monte Carlo Method

An Improved Acceptance Procedure for the Hybrid Monte Carlo Algorithm

A Multiscale Hybrid Method

"Parallel-Tempering"-Assisted Hybrid Monte Carlo Algorithm For Bayesian Inference In Dynamical Systems

Issues in designing hybrid algorithms

Hybrid Monte Carlo-Methods in Credit Risk Management

An irreversible Markov-chain Monte Carlo method with skew detailed balance conditions

A Hybrid Monte Carlo, Discontinuous Galerkin method for linear kinetic transport equations

Analysis of Hybrid MC/Deterministic Methods for Transport Problems Based on Low-Order Equations Discretized by Finite Volume Schemes

A hybrid method of improving source convergence implemented in RMC

Hybridization and postprocessing in finite element exterior calculus

A Hybrid Stochastic Gradient Hamiltonian Monte Carlo Method

Non-Hermitian Polynomial Hybrid Monte Carlo

Overcoming Ergodicity Problems of the Hybrid Monte Carlo Method using Radial Updates

Optimized Monte Carlo Methods

An Efficient Method for Treating Conditional Monte Carlo Simulation

Hybrid Auxiliary Field Quantum Monte Carlo for Molecular Systems

Accelerating the Hybrid Monte Carlo algorithm with ILU preconditioning

The Geometry of Hamiltonian Monte Carlo

Fluid Solver Independent Hybrid Methods for Multiscale Kinetic equations

Hybrid Monte Carlo estimators for multilayer transport problems