ABSTRACT
The generalized linear model (GLIM) was defined in Section 2.3 as a univariate and one-parameter GRM with statistically independent components of the response vector such that transformed means of the components depend linearly on the regression coefficients. “Classically” this is presented using the following definitions.
