Optimal Scaling for Locally Balanced Proposals in Discrete Spaces
This provides a foundational method for improving sampling efficiency in discrete spaces, applicable to areas like training deep energy-based models.
The paper tackles the problem of optimal scaling for Metropolis-Hastings algorithms in discrete spaces, establishing for the first time that efficiency can be characterized by an asymptotic acceptance rate independent of the target distribution, with optimal rates of 0.574 for locally balanced proposals and 0.234 for random walk Metropolis, showing LBP is asymptotically O(N^(2/3)) more efficient than RWM.
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.