|
|||||
|
|
Posterior model probabilities via path-based pairwise priors |
|
| Authors: | James O Berger German Molina† |
| Institution: | Duke University and Statistical and Applied Mathematical Sciences Institute, P.O. Box 14006, RTP, Durham, NC 27709, USA;and Credit Suisse First Boston, Fixed Income Research, One Cabot Square, London, E14 4QJ, UK |
| Abstract: | We focus on Bayesian model selection for the variable selection problem in large model spaces. The challenge is to search the huge model space adequately, while accurately approximating model posterior probabilities for the visited models. The issue of choice of prior distributions for the visited models is also important. |
| Keywords: | Pairwise model comparisons model selection inclusion probabilities search in model space |