Efficient Learning of Random Forest Classifier using Disjoint Partitioning Approach

Random Forest is an Ensemble Supervised Machine Learning technique. Research work in the area of Random Forest aims at either improving accuracy or improving performance. In this paper we are presenting our research towards improvement in learning time of Random Forest by proposing a new approach called Disjoint Partitioning. In this approach, we are using disjoint partitions of training dataset to train individual base decision trees. This helps in creating diversity in base decision trees. Also different subsets of attributes are used at each node of decision tree to increase diversity. This approach generates Random Forest classifier which is trained efficiently and gives classification accuracy comparable to the original Random Forest approach. Index Terms Random Forest, Classification, Decision Tree, Disjoint Partitioning, Learning