Chapter Preview. This chapter describes a unifying framework for the Part I linear model and the binary and count models in Chapters 11 and 12. Generalized linear models, often known by the acronym GLM, represent an important class of nonlinear regression models that have found extensive use in actuarial practice. This unifying framework not only encompasses many models we have seen but also provides a platform for new ones, including gamma regressions for fat-tailed data and “Tweedie” distributions for two-part data.