Boosting additive models using component-wise P-Splines

An efficient approximation of L2 Boosting with component-wise smoothing splines is considered. Smoothing spline base-learners are replaced by P-spline base-learners, which yield similar prediction errors but are more advantageous from a computational point of view. A detailed analysis of the effect of various P-spline hyper-parameters on the boosting fit is given. In addition, a new theoretical result on the relationship between the boosting stopping iteration and the step length factor used for shrinking the boosting estimates is derived.