Binary Response: Link Functions

This lecture covers the concept of binary response in statistics, focusing on the Bernoulli distribution and the linear link function. It explains how to interpret the response variable in terms of a hidden variable, and the different link functions like logit and probit. The lecture also delves into logistic regression for binary data, discussing the common choice of the logit link function and the likelihood estimation based on independent responses. Additionally, it explores the deviances for logistic regression models and the model selection process using information criteria like AIC and BIC.