BIOS 7202 - Generalized Linear Models

[3 Credits]

Three hours of lecture per week. Study of parametric models in the exponential family of distributions including the normal, binomial, Poisson, and gamma. Parameter estimation with iterative re-weighted least squares and quasi-likelihood methods. Modeling of correlated data or data with non-constant variance via mixed models (e.g., GLIMMIX). In-depth coverage of generalized estimating equations (GEE1 and GEE2) and quadratic estimating equations (QEE). Applications with be presented from a variety of settings such as the basic sciences, medicine, dental, and public health. Prerequisite: BIOS 6202, BIOS 6206, or permission of the instructor.