A Survey on Feature Selection Methods for Imbalanced Datasets

Class imbalance problem is one of the greatest challenges in machine learning and data mining researches, which has acquired significant research interest from academics, industries and research teams in recent years. Researchers have proposed many techniques to handle the class imbalance problem, including resampling, new algorithms, and feature selection. The class imbalance problem is even more severe when the dimensionality is high. Both sampling techniques and algorithmic methods may not work well for high dimensional class imbalance problems. Feature or variable selection is a preprocessing technique commonly used for high-dimensional data. The key issue in the feature selection is finding the highly relevant features in the feature sets that allow a classifier to reach optimal performance. This paper presents a survey on feature selection methods for imbalanced datasets.