Sampling From Multiscale Densities With Delayed Rejection Generalized Hamiltonian Monte Carlo

TL;DR

The paper introduces DR-G-HMC, a dynamic step size Hamiltonian Monte Carlo method that efficiently samples from multiscale densities by adaptively adjusting step sizes during the sampling process.

Contribution

It proposes a novel DR-G-HMC algorithm that improves sampling efficiency for hierarchical models with multiscale geometries by employing sequential proposals with decreasing step sizes.

Findings

Successfully samples from multiscale densities

Competitive with No-U-Turn sampler in experiments

Robust to tuning parameters

Abstract

Hamiltonian Monte Carlo (HMC) is the mainstay of applied Bayesian inference for differentiable models. However, HMC still struggles to sample from hierarchical models that induce densities with multiscale geometry: a large step size is needed to efficiently explore low curvature regions while a small step size is needed to accurately explore high curvature regions. We introduce the delayed rejection generalized HMC (DR-G-HMC) sampler that overcomes this challenge by employing dynamic step size selection, inspired by differential equation solvers. In generalized HMC, each iteration does a single leapfrog step. DR-G-HMC sequentially makes proposals with geometrically decreasing step sizes upon rejection of earlier proposals. This simulates Hamiltonian dynamics that can adjust its step size along a (stochastic) Hamiltonian trajectory to deal with regions of high curvature. DR-G-HMC makes generalized HMC competitive by decreasing the number of rejections which otherwise cause inefficient backtracking and prevents directed movement. We present experiments to demonstrate that DR-G-HMC (1) correctly samples from multiscale densities, (2) makes generalized HMC methods competitive with the state of the art No-U-Turn sampler, and (3) is robust to tuning parameters.