☆ 4.6 Article

Joint estimation of Ising model parameters with Hamiltonian constraint

PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS (2024)

Statistically Valid Variational Bayes Algorithm for Ising Model Parameter Estimation

Minwoo Kim et al.

Summary: This article introduces the origin of Ising models in statistical physics and their wide usage in modeling spatial data and computer vision problems. The statistical inference of this model remains challenging due to the intractable nature of the normalizing constant in the likelihood. A pseudo-likelihood approach is used to study the Bayesian estimation of Ising model parameters, and a computationally efficient variational Bayes procedure is developed. The efficacy of mean-field Gaussian and bivariate Gaussian families for inference of Ising model parameters is validated through extensive simulation studies.

JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS (2023)

添加到收藏夹

Article Biochemical Research Methods