Bayesian Model Selection and Forecast Averaging

Jakub Bijak ()
Additional contact information
Jakub Bijak: School of Social Sciences, Centre for Population Change and S3RI, University of Southampton

Chapter Chapter 5 in Forecasting International Migration in Europe: A Bayesian View, 2011, pp 91-116 from Springer

Abstract: Abstract This chapter presents the first out of four perspectives proposed for designing a modelling framework for Bayesian forecasting of international migration. In particular, it explores the Bayesian model selection and forecast averaging techniques, based on the posterior odds criterion. Theoretical foundations are laid in Section 5.1, together with a framework proposed for migration predictions on the basis of simple, non-nested stochastic processes. The computational details of the Carlin–Chib procedure used for model selection are also provided. Section 5.2 presents empirical results of forecasts of international migration among selected European countries for 2005–2015, yielded by various models, with focuses on individual and averaged forecasts from simple stochastic processes.

Keywords: Posterior Probability; Analytic Hierarchy Process; Predictive Distribution; Emigration Rate; High Posterior Density (search for similar items in EconPapers)
Date: 2011
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:ssdmcp:978-90-481-8897-0_5

Ordering information: This item can be ordered from
http://www.springer.com/9789048188970

DOI: 10.1007/978-90-481-8897-0_5

Access Statistics for this chapter

More chapters in The Springer Series on Demographic Methods and Population Analysis from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().