四川大学学报（自然科学版）2024，Vol.61Issue(1)：25-34,10.DOI:10.19907/j.0490-6756.2024.011004
用Velocity Verlet积分器改进HMC抽样方法
Improving the HMC sampling method by Velocity Verlet integrator
摘要
Abstract
Hamilton Monte Carlo(HMC)method is a fast sampling method.To sampling a Hamiltonian equation,the HMC method uses the Leapfrog integrator.As a result,unsynchronized values of position and momentum variables of the equation are possibly generated during the iterative process and resulting estimation error can seriously damage the sampling efficiency and stability of sampling result.To solve this problem,here we propose an improved HMC(IHMC)method by replacing the Leapfrog integrator in HMC method with Velocity Verlet integrator.The IHMC method calculates the variables simultane-ously in each iteration.Two numerical examples are implemented to check the performance of the IHMC method,one is the problem of estimating the parameters of realized and stochastic volatility(RASV)model with asymmetric effects,the other is the problem of sampling the variance Gamma distribution.It is shown that the IHMC method has higher sampling efficiency and more stable sampling result com-pared with the HMC method.关键词
HMC方法/Velocity Verlet积分器/RASV模型/方差伽马分布Key words
Hamiltonian Monte Carlo method/Velocity Verlet integrator/RASV model/Variance Gam-ma distribution分类
数理科学引用本文复制引用
李婉荧,唐亚勇..用Velocity Verlet积分器改进HMC抽样方法[J].四川大学学报（自然科学版）,2024,61(1):25-34,10.
