Efficient simulation of Bayesian logistic regression models

Holmes, C. and Knorr-Held, Leonhard (2003): Efficient simulation of Bayesian logistic regression models. Collaborative Research Center 386, Discussion Paper 306

Abstract

In this paper we highlight a data augmentation approach to inference in the Bayesian logistic regression model. We demonstrate that the resulting conditional likelihood of the regression coefficients is multivariate normal, equivalent to a standard Bayesian linear regression, which allows for efficient simulation using a block Gibbs sampler. We illustrate that the method is particularly suited to problems in covariate set uncertainty and random effects models.

• Item Type: Paper (Research Paper)
• Collections: Mathematics, Computer Science and Statistics > Statistics > Collaborative Research Center 386; Special Research Fields > Special Research Field 386
• Subjects: 500 Science > 510 Mathematics