Ioannis Ntzoufras

Publications Page


 

Bayesian Analysis of the Normal Regression Model

 

Ioannis Ntzoufras

Department of Statistics,

Athens University of Economics and Business,

Athens, GREECE;

e-mail: ntzoufras@aueb.gr

 

Chapter 3 in  Böcker, (ed.) (2010). Rethinking Risk Measurement and Reporting: Uncertainty, Bayesian Analysis and Expert Judgement  Risk Books ISBN-10: 1-906348-40-5, ISBN-13:  978-1-906348-40-3, pp. 69-106.

 

SYNOPSIS

In this chapter, we implement Bayesian methods in regression models which are an essential tool in modern statistical science. They can be used for both  interpretation of social or economic phenomena and prediction of future outcomes which is of major interest in risk analysis.  Possible prior specifications are described in detail. Posterior inference is illustrated focusing on the conjugate case. Bayesian variable selection methods for the conjugate case are also illustrated  while more advanced topics such as variable selection using MCMC and evaluation of the structural assumptions  is briefly discussed accompanied with references for further reading.  The chapter closes with a short discussion and conclusion.

 Download:



Back to Research PageAll Contents Copyright.
Last revised: 12/06/2014