A Random Forest is an ensemble learning method which implements multiple decision trees during training. It predicts by using a combination rule on the outputs of individual decision trees.
See Breiman2001 for a detailed introduction.
CDenseFeatures (here 64 bit floats aka RealFeatures) and CMulticlassLabels are created from training and test data file
random_forest.sg:create_features
Combination rules to be used for prediction are derived form the CCombinationRule class. Here we create a CMajorityVote class to be used as a combination rule.
random_forest.sg:create_combination_rule
Next an instance of CRandomForest is created. The parameters provided are the number of attributes to be chosen randomly to select from and the number of trees.
random_forest.sg:create_instance
Then we run the train random forest and apply it to test data, which here gives CMulticlassLabels.
random_forest.sg:train_and_apply
We can evaluate test performance via e.g. CMulticlassAccuracy as well as get the "out of bag error".
random_forest.sg:evaluate_accuracy
Random_forest
Out-of-bag_error
../../references.bib