STAT 5067

This course builds upon the material of Statistics 4066/5066. After a review of probability theory, statistical theory for the major methods of estimation will be rigorously developed. Topics include statistical consistency, limiting distributions of estimators, limiting distributions for testing in likelihood settings and transformations of confidence regions. Asymptotic optimality for point estimation, testing and confidence regions will be defined and optimality results will be established for likelihood methods. Laplace approximation will be used to investigate the properties of Bayesian methods and to derive the BIC model selection criterion.