Bayesian Model Comparison: Conference and Research Volume

Model uncertainty is a key component of statistical data analysis and an integral part in the inferential process. Because theory typically implies a range of possible competing empirical specifications, accounting for model uncertainty is crucial for understanding the processes under investigation, and is a necessary step in interpreting model parameters and performing predictions. Bayesian model comparison and model averaging is an active research area that continues to generate new ideas and innovative approaches.

The aim of this call for papers is to produce a research volume that examines key aspects of modern Bayesian research on model comparison and model averaging. The volume will address important challenges in this area with the goal of improving theoretical foundations and practical implementation.  

Possible topics include, but are not limited to:

Approaches for evaluating marginal likelihoods and Bayes factors Comparative studies of alternative methods Computational issues in model comparison Variable selection methods, Bayesian LASSO Comparisons of semi-parametric and non-parametric models Approximate methods, asymptotic behavior, information criteria The importance of prior assumptions Forecasting under model uncertainty, model averaging Applications in economics, statistics, and the social sciences Selected papers will appear in Advances in Econometrics, Volume 34. The volume will be edited by Dale Poirier and Ivan Jeliazkov. Please e-mail extended abstracts or complete papers no later than January 10, 2014 to: 

dpoirier@uci.edu and ivan@uci.edu.

A research conference for contributors will be held at the University of California, Irvine, February 22-23, 2014. Review of manuscripts will commence soon after the conference and accepted articles will appear in print in Fall 2014/Winter 2015.

Advances in Econometrics is a research annual whose editorial policy is to publish original articles that contain enough details so that economists and econometricians who are not experts in the topics will find them accessible and useful in their research. Authors should be able to provide, upon request, computer programs and data used in their articles. For more information on the Advances in Econometrics series and the contents of previous volumes, see http://faculty.smu.edu/millimet/AiE.html.