Abstract

Regression models are typically constructed to model the mean and dispersion/precision of a
distribution. However, the density or probability mass function of several distributions is not
indexed by the mean and dispersion/precision parameters. In this context, this work
provides a collection of regression models considering new parameterizations in terms of
the mean and dispersion/precision parameters. The main advantage of our new
parametrizations is the straightforward interpretation of the regression
coefficients in terms of the expectation and dispersion, as usual in the context of
generalized linear models. The maximum likelihood method is used to estimate the model parameters.