Generalized Linear Models

This course develops generalized linear models using likelihood-based methods, canonical links, and iteratively re-weighted least squares. It covers probit, logistic, complementary log-log, Poisson, ordinal logistic, and multinomial logistic regression.

Syllabus

Generalized Linear Models: Background.
Generalized Linear Models: Canonical Link Function.
Generalized Linear Models: Likelihood, Score, and Fisher Information.
GLM: Iteratively Re-weighted Least Squares for a General Link Function.
Generalized Linear Models: Probit Regression (part 1).
Generalized Linear Models: Probit Regression (part 2).
Generalized Linear Models: Logistic "Logit" Regression (part 1).
Generalized Linear Models: Logistic "Logit" Regression (part 2).
Generalized Linear Models: Logistic "Logit" Regression (part 2).
Generalized Linear Models: Complementary Log Log Regression (part 1).
Generalized Linear Models: Complementary Log Log Regression (part 2).
Generalized Linear Models: Complementary Log Log Regression (part 2).
Generalized Linear Models: Poisson Regression with Canonical Link (part 1).
Generalized Linear Models: Poisson Regression with Canonical Link (part 2).
Ordinal Logistic Regression (Proportional Odds Model).
Multinomial Logistic Regression.