Association Rule Based Undersampling Technique for Addressing Imbalanced Classification
Zahid Ahmed, Sufal Das · 2024
Imbalanced classification is a crucial challenge in the machine learning domain. Due to unequal instances in different classes, the performance of traditional classifiers may decrease. This paper proposes a novel undersampling approach based on association rules known as Association Rule- Based Undersampling (ARU) to efficiently handle class imbalances and preserve the essential characteristics of the original dataset. This approach incorporates a weighted concept to identify appropriate majority-class instances that are important to eliminate. It can carefully identify identical and excessive instances and eliminate them from the majority class. Extensive experiments have been conducted using diverse benchmark datasets to compare the performance with various state-of-the-art methods. The results demonstrate that the proposed approach outperforms existing approaches by achieving better classification results and can efficiently address the imbalanced problem.