Generalized Linear Models II: GLM Extensions

This lecture covers advanced topics in Generalized Linear Models (GLM), focusing on link functions, error distributions, and model interpretation. It delves into Bernoulli and Binomial observations, choice of link functions, and the impact of sparseness on model interpretability. The lecture also explores separation issues, jittered residuals, and contingency tables in GLM. Additionally, it discusses nonparametric relationships with covariates, kernel smoothing techniques, and the challenges of fitting models with perfect separation. The instructor provides insights on multinomial loglikelihood, count data analysis, and the implications of separation in GLM.