This page contains details about the GLM course (IM60208).
Semesters Taught: Spring 2022–23, 2023–24, 2024–25
| Module | Title | Topics Covered |
|---|---|---|
| Module 1 | Basics of Estimation and Univariate Discrete Data Modelling | Probability review; Point estimation (Method of Moments, MLE); Properties of estimators; Likelihood and log-likelihood; Bernoulli, Binomial, Poisson models; Estimation and inference |
| Module 2 | Bivariate Categorical Data Modelling | Contingency tables; Marginal and conditional distributions; Measures of association; Independence; Chi-square tests; Fisher’s exact test |
| Module 3 | Linear Regression | Simple and multiple regression; Model assumptions; Least squares estimation; Inference for parameters; Diagnostics and limitations |
| Module 4 | Generalized Linear Models and Components | Motivation for GLMs; Exponential family; Random, systematic, and link components; Canonical links; Likelihood inference; IRLS |
| Module 5 | Modelling of Binary Responses and Extensions | Binary response models; Logistic regression; Probit and complementary log-log links; Odds ratio interpretation; Diagnostics; Multinomial and ordinal models (overview) |
| Module 6 | Count Responses and Loglinear Models | Poisson regression; Exposure and offsets; Loglinear models for contingency tables; Parameter interpretation; Applications |
| Module 7 | Survival Analysis | Time-to-event data; Censoring; Survival and hazard functions; Kaplan–Meier estimator; Cox proportional hazards model |
| Module 8 | Overdispersion and Quasi-likelihood Estimation | Sources of overdispersion; Diagnostics; Quasi-likelihood framework; Quasi-Poisson and binomial models; Robust standard errors |
| Module 9 (Optional) | GLMs and Bayesian Statistics | Bayesian approach to GLMs; Prior distributions; Posterior inference; Bayesian logistic and Poisson regression; Introduction to MCMC |
Lecture notes, problem sets, etc. are provided. Check the course moodle page for more recent updates.