Exercise

Tree-based AdaBoost regression

AdaBoost models are usually built with decision trees as the base estimators. Let's give this a try now and see if model performance improves even further.

We'll use twelve estimators as before to have a fair comparison. There's no need to instantiate the decision tree as it is the base estimator by default.

Instructions

100 XP
  • Build and fit an AdaBoostRegressor using 12 estimators. You do not have to specify a base estimator.
  • Calculate the predictions on the test set.