By: BIOLAB, Dec 8, 2011

Random forest switches to Simple tree learner by default

Random forest classifiers now use Orange.classification.tree.SimpleTreeLearnerby default, which considerably shortens their construction times. Using a random forest classifier is easy. import Orange iris = Orange.data.Table('iris') forest = Orange.ensemble.forest.RandomForestLearner(iris, trees=200) for instance in iris: print forest(instance), instance.get_class() The example above loads the iris dataset and trains a random forest classifier with 200 trees. The classifier is then used to label all training examples, printing its prediction alongside the actual class value.