Random Forest

An Ensemble Method consisting only of decision trees where each tree has been generated using random attribute selections at each node.

Each tree will give a vote and the class with the most votes will be assigned to the sample.

There are two major methods to construct random forests:

Random Forests are comparable in Accuracy to AdaBoost but are more robust to errors and Outliers. Also they are faster than most or all Bagging and Boosting methods.