This is similar to the previous recipes, but we use Random Forest Trees to solve a regression problem (continuous). The following parameter is used to direct the algorithm to apply regression rather than classification. We again limit the number of classes to two:
val impurity = "variance" // USE variance for regression