The research topic is Bayesian nonparametric inference. Foundational questions in coherence and admissibility will be investigated. Nonparametric prediction and distribution estimation will be studied by constructing new models and extending existing ones. Asymptotic expansions will be developed leading to computable approximations to the posterior distributions of interesting population functionals. Bayesian nonparametric ideas and procedures will be applied to specific problems in survival analysis, treatment allocation and comparison, and the analysis of histogram data. In cases that exact Bayesian nonparametric procedures turn out to be impossible or difficult to derive, alternative inferential procedures with a quasi-Bayes flavor will be explored.
