Review of Random Forest Classification Techniques to Resolve Data Imbalance

In this current age, numerous ranges of real word applications with imbalanced dataset is one of the foremost focal point of researcher's inattention. There is the enormous increment of data generation and imbalance within dataset. Processing and knowledge extraction of huge amount of imbalanced data becomes a challenge related with space and time necessities. Generally there is a list of an assortment of factual humanity applications which deals with unequal data sample division in to number of classes. Due to this division of data either of class goes into majority or minority with comparably less data count. This outnumbering of data sample in either of one class directs towards the handling of minority class and target on remarkable reduction in error rate. The standard learning methods do not directly focus on this type of classes. Random Forest Classification (RFC) is an ensemble approach that utilizes a number of classifiers to work together in order to identify the class label for unlabeled instances. This approach has proved its high accuracy and superiority with imbalanced datasets. This classifier provides various techniques to resolve class imbalance problem. This paper summarizes, the literature survey from 2000 to 2016 of various techniques related to RFC to resolve class imbalance. Specifically Weighted Random Forest (WRF), Balanced Random Forest (BRF), Sampling (Under Sampling (US)) and Down Sampling (DS), Cost Sensitive Methods have been adapted more to till date. The limitation of this numerous literature is researchers can focus on dynamic integration techniques to resolve class imbalance and increase robustness and versatility of classification.