Slide 33: Models and their selection - GAM modeling
GAM modelling is based on the simpler form of linear regression modelling. For linear regression there is a dependent variable Y and predictor variables X1 ... Xp such that the dependent variable is equivalent to the sum of the products of the linear function Bj multiplied by the independent predictor variable functions, with a small correction factor.
Additive models replace the linear function Bj with a smoothed non-linear function fj, as outlined in the second equation. Since the dependent variable (presence/absence) is binomial in nature, the nonlinear function must be of the nonlinear transformation family (LOGIT), in order to output a probabilistic outcome.