Generalized additive models
A GAM (Hastie and Tibshirani 1986) replaces the linear predictor in a GLM by additive smooth functions of the explanatory variables.
\[g(\mathbb E[Y \mid X = x]) = \beta_0 + s_1(x_1) + s_2(x_2) + \dots + s_p(x_p)\]
A GAM (Hastie and Tibshirani 1986) replaces the linear predictor in a GLM by additive smooth functions of the explanatory variables.
\[g(\mathbb E[Y \mid X = x]) = \beta_0 + s_1(x_1) + s_2(x_2) + \dots + s_p(x_p)\]