English

High-dimensional Inference and FDR Control for Simulated Markov Random Fields

Machine Learning 2024-01-23 v3 Machine Learning Statistics Theory Statistics Theory

Abstract

Identifying important features linked to a response variable is a fundamental task in various scientific domains. This article explores statistical inference for simulated Markov random fields in high-dimensional settings. We introduce a methodology based on Markov Chain Monte Carlo Maximum Likelihood Estimation (MCMC-MLE) with Elastic-net regularization. Under mild conditions on the MCMC method, our penalized MCMC-MLE method achieves 1\ell_{1}-consistency. We propose a decorrelated score test, establishing both its asymptotic normality and that of a one-step estimator, along with the associated confidence interval. Furthermore, we construct two false discovery rate control procedures via the asymptotic behaviors for both p-values and e-values. Comprehensive numerical simulations confirm the theoretical validity of the proposed methods.

Keywords

Cite

@article{arxiv.2202.05612,
  title  = {High-dimensional Inference and FDR Control for Simulated Markov Random Fields},
  author = {Haoyu Wei and Xiaoyu Lei and Yixin Han and Huiming Zhang},
  journal= {arXiv preprint arXiv:2202.05612},
  year   = {2024}
}
R2 v1 2026-06-24T09:31:59.427Z