(i) To develop optimal sequential testing and estimation procedures for biomedical data in general and censored survival data in particular. (ii) To find efficient design of clinical trials for comparing two treatments, incorporating the scientific, economic and ethical considerations. (iii) To compare the linear and general empirical Bayes approaches to estimating many parameters (means, variances, etc.) and to construct adaptive empirical Bayes estimators that will combine the best features of both. To extend empirical Bayes methods to completely nonparametric problems. To construct tests based on empirical Bayes prediction intervals for the efficacy of a medical treatment without the necessity of using a control group. (iv) To develop and apply pattern recognition and nonparametric classification techniques for certain clinical problems. (v) To develop stochastic approximation and other techniques for bioassay and dosage determination.
