Advanced Regression for Independent Data

3 credits · Graduate School

This course is the first in a two‐course sequence on advanced regression methodology. The first half of the course covers ordinary and weighted least squares, regression-based limit theorems, analysis of variance (ANOVA), hypothesis testing, confidence and prediction regions, model misspecification, and diagnostics. The second half of the course covers generalized linear models (e.g., binomial, Poisson, and gamma regression) from the likelihood and Z-estimation frameworks, overdispersion and quasi-likelihood, hypothesis testing, diagnostics, and negative binomial regression. Gauss-Markov optimality is a central theme of this course. As time permits, advanced topics may be featured such as regularization, nonlinear least squares, cumulative probability models, receiver operating characteristic regression, and matched pair designs. This course emphasizes methodology, computational exploration, and theoretical underpinnings in roughly equal parts. This course is open to biostatistics PhD students only. The course is co-located with and serves as repeat credit for BIOS 7345. Prerequisites: BIOS 6312 and BIOS 6342. [3]