Using R to run a classifier

Weka’s MLR classifier includes many of the learning algorithms that are available in the R environment. Choosing the MLRClassifier in the Explorer’s Classify panel gives access to 75 classification methods and 60 regression methods. Behind the scenes, Weka transfers the data into the R environment, builds the classifier or regressor there, feeds in the test data, and extracts the predictions from the R environment and returns them to Weka. In the video Eibe runs R’s rpart method, which builds a classification tree; and rferns, which generates an ensemble of ferns – a restricted form of decision tree. Detailed instructions are given in the accompanying download (these slides do not appear in the video itself).