Martin Tanner Professor of Statistics
My research interests include Markov chain Monte Carlo methods for Bayesian and frequentist inference, nonparametric estimation of the hazard function for right-censored and interval-censored data, applications of multiple imputation to censored regression data, models and measures of interrater agreement/ disagreement, and mathematical models of carcinogenesis. My recent work considers the use of Bayesian inference in mixtures-of-experts and hierarchical mixtures-of-experts neural network architectures with applications to speech recognition, breast cancer diagnosis, financial exchange rate data, and global meteorological information.
Mixtures of marginal models (with Ori Rosen and Wenxin Jiang) Biometrika 87 (2000), 391-404.
Binomial-beta hierarchical models for ecological inference (with Gary King and Ori Rosen) Sociological Methods and Research 28 (1999), 61-90.
Bayesian inference in mixtures-of-experts and hierarchical mixtures-of-experts models with an application to speech recognition (with F. Peng and R. Jacobs). Journal of the American Statistical Association 91 (1996), 953-960.
Tools for Statistical Inference: Methods for the Exploration of Posterior Distributions and Likelihood Functions (1996), Springer-Verlag, 3rd ed.
The calculation of posterior distributions by data augmentation (with discussion) (with W H. Wong). Journal of the American Statistical Association 82 (1987), 528-50.