Polish Statistical Association
Central Statistical Office of Poland
Subject: Economics, Statistics & Probability
ISSN: 1234-7655
eISSN: 2450-0291
SEARCH WITHIN CONTENT
Keywords : Bayesian approach, regression models, a priori information, MCMC
Citation Information : Statistics in Transition New Series. Volume 17, Issue 4, Pages 763-780, DOI: https://doi.org/10.21307/stattrans-2016-051
License : (CC BY 4.0)
Published Online: 06-July-2017
In this study the benefits arising from the use of the Bayesian approach to predictive modelling will be outlined and exemplified by a linear regression model and a logistic regression model. The impact of informative and non-informative prior on model accuracy will be examined and compared. The data from the Central Statistical Office of Poland describing unemployment in individual districts in Poland will be used. Markov Chain Monte Carlo methods (MCMC) will be employed in modelling.