RHSOFS: Feature Selection Using the Rock Hyrax Swarm Optimization Algorithm for Credit Card Fraud Detection System.

Area of Science:

• Computer Science
• Artificial Intelligence
• Machine Learning

Background:

• Credit card fraud detection is challenging due to high-dimensional and imbalanced datasets.
• Effective feature selection is crucial for enhancing classification performance in fraud detection.

Purpose of the Study:

• To propose a novel feature selection (FS) approach, Rock Hyrax Swarm Optimization Feature Selection (RHSOFS), for credit card fraud detection.
• To improve the identification of fraudulent transactions using supervised machine learning techniques.

Main Methods:

• Developed RHSOFS, a metaheuristic algorithm inspired by rock hyrax swarm behavior.
• Applied RHSOFS to select optimal relevant features from high-dimensional datasets.
• Implemented supervised machine learning for fraud transaction identification.

Main Results:

• RHSOFS demonstrated superior performance in feature selection compared to Differential Evolutionary Feature Selection (DEFS), Genetic Algorithm Feature Selection (GAFS), Particle Swarm Optimization Feature Selection (PSOFS), and Ant Colony Optimization Feature Selection (ACOFS).
• Experimental results confirmed the effectiveness of the proposed RHSOFS approach.
• Statistical tests validated the significance of the RHSOFS model.

Conclusions:

• The proposed RHSOFS is an effective metaheuristic approach for feature selection in credit card fraud detection.
• RHSOFS enhances the accuracy of identifying fraudulent transactions by selecting optimal feature subsets.
• This method offers a significant advancement over existing feature selection techniques in this domain.