Optimal Scaling for Locally Balanced Proposals in Discrete Spaces

TL;DR

This paper extends the understanding of optimal scaling in Metropolis-Hastings algorithms to discrete spaces, identifying optimal acceptance rates and demonstrating improved efficiency and adaptive tuning methods.

Contribution

It establishes the asymptotic acceptance rate for discrete M-H algorithms and compares the efficiency of locally balanced proposals to RWM, introducing adaptive tuning in discrete spaces.

Findings

Optimal acceptance rate for LBP is 0.574.

Optimal acceptance rate for RWM is 0.234.

LBP is asymptotically more efficient than RWM by a factor of O(N^{2/3}).

Abstract

Optimal scaling has been well studied for Metropolis-Hastings (M-H) algorithms in continuous spaces, but a similar understanding has been lacking in discrete spaces. Recently, a family of locally balanced proposals (LBP) for discrete spaces has been proved to be asymptotically optimal, but the question of optimal scaling has remained open. In this paper, we establish, for the first time, that the efficiency of M-H in discrete spaces can also be characterized by an asymptotic acceptance rate that is independent of the target distribution. Moreover, we verify, both theoretically and empirically, that the optimal acceptance rates for LBP and random walk Metropolis (RWM) are $0.574$ and $0.234$ respectively. These results also help establish that LBP is asymptotically $O(N^\frac{2}{3})$ more efficient than RWM with respect to model dimension $N$. Knowledge of the optimal acceptance rate allows one to automatically tune the neighborhood size of a proposal distribution in a discrete space, directly analogous to step-size control in continuous spaces. We demonstrate empirically that such adaptive M-H sampling can robustly improve sampling in a variety of target distributions in discrete spaces, including training deep energy based models.