In this video, we go over a brief introduction to Multivariate Adaptive Regression Splines (MARS). Then, we use MARS to predict a continuous response variable, with the Boston housing dataset. We do this using the "earth" package in R, and go over how to use cross-validation and parameter tuning (through a grid search) using the "caret" package.

To view the code from this tutorial, go here: https://github.com/melissavanbussel/Y...