Learning Mixtures of Ising Models using Pseudolikelihood
This addresses a specific statistical learning problem for researchers in statistical physics or machine learning, but appears incremental as it extends an existing method to a mixture setting.
The paper tackled the problem of learning parameters for mixtures of Ising models by deriving a pseudolikelihood method, demonstrating its performance on synthetic and real data for Ising and Potts models.
Maximum pseudolikelihood method has been among the most important methods for learning parameters of statistical physics models, such as Ising models. In this paper, we study how pseudolikelihood can be derived for learning parameters of a mixture of Ising models. The performance of the proposed approach is demonstrated for Ising and Potts models on both synthetic and real data.