Course Description
Course Name
Decision Theory and GLM
Session: VCPF3125
Hours & Credits
36 Host University Units
Prerequisites & Language Level
Course entry requirements: STA2004F and STA2005S. MAM2000W is strongly recommended (linear algebra and advanced calculus modules).
Taught In English
- There is no language prerequisite for courses at this language level.
Overview
Course outline:
This course forms part of the third year major in Mathematical Statistics. It consists of two modules: The Generalised Linear Models module introduces students to the theory and application of fitting linear models to different types of response variables with different underlying distributions. The Decision and Risk Theory module is an introduction to the structure of decision making under uncertainty. The content of the modules are as follows:
Generalized linear modules: Topics covered include: the exponential family of distributions, the GLM formulation, estimation and inference, models for continuous responses with skew distributions, logistic regression, Poisson regression and loglinear models.
Decision theory: Topics covered include: game theory and non probabilistic decision criteria; probabilistic decision criteria; expected value and utility; use of Bayes? theorem; value of information; Bayesian statistical analysis for Bernoulli and normal sampling; empirical Bayes and credibility theory; loss and extreme value distributions; Monte Carlo method.
This course forms part of the third year major in Mathematical Statistics. It consists of two modules: The Generalised Linear Models module introduces students to the theory and application of fitting linear models to different types of response variables with different underlying distributions. The Decision and Risk Theory module is an introduction to the structure of decision making under uncertainty. The content of the modules are as follows:
Generalized linear modules: Topics covered include: the exponential family of distributions, the GLM formulation, estimation and inference, models for continuous responses with skew distributions, logistic regression, Poisson regression and loglinear models.
Decision theory: Topics covered include: game theory and non probabilistic decision criteria; probabilistic decision criteria; expected value and utility; use of Bayes? theorem; value of information; Bayesian statistical analysis for Bernoulli and normal sampling; empirical Bayes and credibility theory; loss and extreme value distributions; Monte Carlo method.
*Course content subject to change