Watson Wyatt, 2005. — 116 p.
The Practitioner's Guide to Generalized Linear Models is written for the practicing actuary who would like to understand generalized linear models (GLMs) and use them to analyze insurance data. The guide is divided into three sections.
GLMs - theory and intuition
GLMs in practice
Other applications of GLMs
AppendixThe design matrix when variates are used
The exponential family of distributions
The Tweedie distribution
Canonical link functions
Solving for maximum likelihood in the general case of an exponential distribution
Example of solving for maximum likelihood with a gamma error and inverse link function
Data required for a GLM claims analysis
Automated approach for factor categorization
Cramer's V
Benefits of modeling frequency and severity separately rather than using Tweedie GLMs