Random Forest Classifiers :A Survey and Future Research Directions

Random Forest is an ensemble supervised machine learning technique. Machine learning techniques have applications in the area of Data mining. Random Forest has tremendous potential of becoming a popular technique for future classifiers because its performance has been found to be comparable with ensemble techniques bagging and boosting. Hence, an in-depth study of existing work related to Random Forest will help to accelerate research in the field of Machine Learning. This paper presents a systematic survey of work done in Random Forest area. In this process, we derived Taxonomy of Random Forest Classifier which is presented in this paper. We also prepared a Comparison chart of existing Random Forest classifiers on the basis of relevant parameters. The survey results show that there is scope for improvement in accuracy by using different split measures and combining functions; and in performance by dynamically pruning a forest and estimating optimal subset of the forest. There is also scope for evolving other novel ideas for stream data and imbalanced data classification, and for semi-supervised learning. Based on this survey, we finally presented a few future research directions related to Random Forest classifier.