Comparing and calibrating discrepancy measures for Bayesian model selection

Different approaches have been considered in the literature for the problem of Bayesian model selection. Recently, a new method was introduced and analysed in De la Horra (2008) by minimizing the posterior expected discrepancy between the set of data and the Bayesian model, where the chi-square disc...

Descripción completa

Detalles Bibliográficos
Autores: Horra, Julián de la, Rodríguez Bernal, María Teresa
Tipo de recurso: artículo
Fecha de publicación:2012
País:España
Institución:Universitat Autònoma de Barcelona
Repositorio:Dipòsit Digital de Documents de la UAB
Idioma:inglés
OAI Identifier:oai:ddd.uab.cat:97727
Acceso en línea:https://ddd.uab.cat/record/97727
Access Level:acceso abierto
Palabra clave:Bayesian model selection
Discrepancy measure
Calibration
Posterior expected discrepancy
Descripción
Sumario:Different approaches have been considered in the literature for the problem of Bayesian model selection. Recently, a new method was introduced and analysed in De la Horra (2008) by minimizing the posterior expected discrepancy between the set of data and the Bayesian model, where the chi-square discrepancy was used. In this article, several discrepancy measures are considered and compared by simulation, and it is obtained that the chi-square discrepancy is reasonable to use. Then, an easy method for calibrating discrepancies is proposed, and the behaviour of this approach is studied on simulated data. Finally, a set of real data is analysed.