English

On approximating the $f$-divergence between two Ising models

Data Structures and Algorithms 2025-09-08 v1 Machine Learning Probability

Abstract

The ff-divergence is a fundamental notion that measures the difference between two distributions. In this paper, we study the problem of approximating the ff-divergence between two Ising models, which is a generalization of recent work on approximating the TV-distance. Given two Ising models ν\nu and μ\mu, which are specified by their interaction matrices and external fields, the problem is to approximate the ff-divergence Df(νμ)D_f(\nu\,\|\,\mu) within an arbitrary relative error e±ε\mathrm{e}^{\pm \varepsilon}. For χα\chi^\alpha-divergence with a constant integer α\alpha, we establish both algorithmic and hardness results. The algorithm works in a parameter regime that matches the hardness result. Our algorithm can be extended to other ff-divergences such as α\alpha-divergence, Kullback-Leibler divergence, R\'enyi divergence, Jensen-Shannon divergence, and squared Hellinger distance.

Keywords

Cite

@article{arxiv.2509.05016,
  title  = {On approximating the $f$-divergence between two Ising models},
  author = {Weiming Feng and Yucheng Fu},
  journal= {arXiv preprint arXiv:2509.05016},
  year   = {2025}
}
R2 v1 2026-07-01T05:22:56.804Z