Introduction

This note outlines a simple system for fitting generalized linear models in Lisp-Stat. Three standard models are implemented:

• Poisson regression models
• Binomial regression models
• Gamma regression models

The next section outlines the basic use of the generalized linear model objects. The third section describes a few functions for handling categorical independent variables. The fourth section gives further details on the structure of the model prototypes, and describes how to define new models and link structures. The final section illustrates several ways of fitting more specialized models, using the Bradley-Terry model as an example.