|
Abstract:
|
We address current methodological research on nonparametric Random Forests, beginning with a brief intellectual history of Random Forests that covers CART, boosting and bagging methods. The primary methods by which researchers can visualize results, the relationships between covariates and responses, and the generalization (out-of-bag prediction) error as the forest is being grown are illustrated. Current research on universal consistency and importance tests in Random Forests is also discussed. Finally, several uses for Random Forests, and some software considerations, are noted.
|
