{"id":"arxiv/2601.01190","url":"https://arxiv.org/abs/2601.01190","tags":["stat.ME","math.PR"],"score":0,"title":"A Modified Bayesian Criterion for Model Selection in Mixed and Hierarchical Frameworks","domain":"stat.ME","source":"arxiv","authors":["Diogenes de Jesus Ramirez","Anderson Melchor Hernandez","Isabel Cristina Ramirez","Luis Raúl Pericchi"],"reviews":0,"subject":"stat.ME","summary":"In this work, we propose a modified Bayesian Information Criterion (BIC) specifically designed for mixture models and hierarchical structures. This criterion incorporates the determinant of the Hessian matrix of the log-likelihood function, thereby refining the classical Bayes Factor by accounting for the curvature of the likelihood surface. Such geometric information introduces a more nuanced penalization for model complexity. The proposed approach improves model selection, particularly under small-sample conditions or in the presence of noise variables. Through theoretical derivations and…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-01-03","reactions":{}}