A Randomized Pairwise Likelihood Method for Complex Statistical Inferences

• 成果类型: Article
• 署名作者: Mazo, Gildas; Karlis, Dimitris; Rau, Andrea
• 署名单位: INRAE; Universite Paris Saclay; Athens University of Economics & Business; INRAE; Universite Paris Saclay; AgroParisTech; Universite de Lille; Universite de Picardie Jules Verne (UPJV); INRAE
• 刊物名称: JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
• ISSN/ISSBN: 0162-1459
• DOI: 10.1080/01621459.2023.2257367
• 发表日期: 2024
• 页码: 2317-2327
• 关键词: variational inference; models; regression

摘要

Pairwise likelihood methods are commonly used for inference in parametric statistical models in cases where the full likelihood is too complex to be used, such as multivariate count data. Although pairwise likelihood methods represent a useful solution to perform inference for intractable likelihoods, several computational challenges remain. The pairwise likelihood function still requires the computation of a sum over all pairs of variables and all observations, which may be prohibitive in high dimensions. Moreover, it may be difficult to calculate confidence intervals of the resulting estimators, as they involve summing all pairs of pairs and all of the four-dimensional marginals. To alleviate these issues, we consider a randomized pairwise likelihood approach, where only summands randomly sampled across observations and pairs are used for the estimation. In addition to the usual tradeoff between statistical and computational efficiency, it is shown that, under a condition on the sampling parameter, this two-way random sampling mechanism makes the individual bivariate likelihood scores become asymptotically independent, allowing more computationally efficient confidence intervals to be constructed. The proposed approach is illustrated in tandem with copula-based models for multivariate count data in simulations, and in real data from a transcriptome study. Supplementary materials for this article are available online.