BIOS 7345 – Linear & Generalized Linear Models/
BIOS 8345 – Advanced Regression for Independent Data

Announcements:  Welcome to BIOS 7345/8345! This will be the course page for Fall 2026. All slides are dated for version control.

Syllabus: A draft Fall 2026 syllabus can be found here.

Lecture notes: Although it is my hope that the materials on this page are broadly helpful, they are not to be copied without permission of the instructor.

# Topic Slides
01 Random vectors and matrices Link
02 Ordinary least squares (OLS; full-rank) Link
03 Hypothesis testing and ANOVA Link
04 Weighted least squares Link
05 Confidence regions and prediction Link
06 Diagnostics Link
07 (Review of) exponential families Link
08 Generalized linear models Link
09 Sandwich and bootstrap methods Link
10 Overdispersion, quasi-likelihood, and optimality Link
11 Hypothesis testing for GLMs Link
12 Diagnostics for GLMs Link
13 Further notes on binary outcomes Link
14 Nonlinear least squares Link
15 Supplemental: Linear algebra Link
16 Supplemental: OLS (rank-deficient) Link
17 Supplemental: Regularization Link
18 Supplemental: Bayesian linear regression Link

Accessibility: Slide decks now include alt text (drafted with AI assistance) for figures in the form of insertions next to each caption. This functionality appears to work best with Adobe.

Problem sets: The course problems can be found here. Problems are due in batches (see schedule below). Certain problems are designated as being required only for students enrolled in BIOS 8345 (all other students would benefit from at least reading the questions and should follow the main points if we spend time discussing solutions in class). Students for whom such questions are not required may elect to complete these problems and turn them in with their homework assignments for a modest amount of extra credit.

# Due date BIOS 7345/8345 BIOS 8345 only
01 09/04 1-4
02 09/11 5-7 7
03 09/18 8-10 10
04 10/02 11-12
05 10/09 13-17 17
06 10/16 18-19
07 10/21 20-22 22
08 11/06 23-24
09 11/13 25-26
10 11/20 27
11 12/04 28-31 31
12 12/11 32-33 33

Data: Data sets and documentation can be found here (you will need the password).