A sequential reduction method for inference in generalized linear mixed\n models

2013/12/06 by Helen Ogden, Ogden, Helen
Decision Sciences · Environmental Science · #Computation (stat.CO) #FOS: Computer and information sciences #Machine Learning (stat.ML) #Methodology (stat.ME) #Optimal Experimental Design Methods #Probabilistic and Robust Engineering Design #Soil Geostatistics and Mapping

paper · pdf · doi:10.48550/arxiv.1312.1903

openalex publication_date 2013/12/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

The likelihood for the parameters of a generalized linear mixed model\ninvolves an integral which may be of very high dimension. Because of this\nintractability, many approximations to the likelihood have been proposed, but\nall can fail when the model is sparse, in that there is only a small amount of\ninformation available on each random effect. The sequential reduction method\ndescribed in this paper exploits the dependence structure of the posterior\ndistribution of the random effects to reduce substantially the cost of finding\nan accurate approximation to the likelihood in models with sparse structure.\n

Related

• Inference for generalized additive mixed models via penalized marginal likelihood
• Conditional Inference for Multivariate Generalised Linear Mixed Models
• Exact MLE for Generalized Linear Mixed Models
• Fixed effects Selection in high dimensional Linear Mixed Models
• Approximate likelihood inference in generalized linear latent variable\n models based on integral dimension reduction