A conceptual model of enhanced undersampling technique

• Main Authors: Zorkeflee, Maisarah, Ku-Mahamud, Ku Ruhana, Mohamed Din, Aniza
• Format: Conference or Workshop Item
• Language: English
• Published: 2014
• Subjects: QA76 Computer software
• Online Access: http://repo.uum.edu.my/13093/1/PID307%20-%20maisarah.pdf; http://repo.uum.edu.my/13093/; http://www.kmice.cms.net.my/
• Institution: Universiti Utara Malaysia

Imbalanced datasets often lead to decrement of classifiers’ performance.Undersampling technique is one of the approaches that is used when dealing with imbalanced datasets problem.This paper discusses on the advantages and disadvantages of several undersampling techniques.An enhanced Distancebased undersampling technique is proposed to balance the imbalanced data that will be used for classification. The fuzzy logic has been integrated in the distance-based undersampling technique to resolve the ambiguity and bias issues.