Research Interests

Bayesian statistical methodology motivated by complex biomedical data and machine learning applications. Ongoing methodologic research focuses on nonparametric Bayes, latent variable methods, variable and covariance selection in high dimensions, density regression, functional data analysis and mixture models. A particular focus has been on developing novel nonparametric Bayes priors for local mixture models, with the goal of improving flexibility, while favoring a sparse formulation of high-dimensional distributions. This research has been motivated by epidemiology studies and machine learning, but has broad applications. Methodologic work is strongly motivated by real world applications, including (but not limited to) reproductive and genetic epidemiology and machine learning.

Recent Publications.

Submitted Manuscripts.

Course Webpage: STA216: Generalized Linear Models

Recent Awards: 2007 Fellow of the American Statistical Association

2007 Spiegelman Award for top biostatistician under 40

2007 Gold Medal for exceptional service from EPA

Latest painting

Go to the Department of Statistical Science -->

2010 Spiegelman Award Call for Nominations.