A Weighted Ensemble Learning Approach Enhanced by the Cuckoo Search Algorithm for Credit Card Fraud Detection
Mohammed Tayebi
Tayebi Mohammed, Faculty of Sciences and Techniques, Hassan First University of Settat, Settat, Morocco.
Manuscript received on 26 February 2023 | Revised Manuscript received on 05 March 2023 | Manuscript Accepted on 15 March 2023 | Manuscript published on 30 March 2023 | PP: 33-35 | Volume-12 Issue-5, April 2023 | Retrieval Number: 100.1/ijitee.D105314040325 | DOI: 10.35940/ijitee.D1053.12050423
Open Access | Editorial and Publishing Policies | Cite | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The prevention and identification of unauthorised credit card transactions play a crucial role in combating credit card fraud, as such transactions can lead to significant financial losses for both cardholders and financial institutions. This article introduces a novel ensemble learning approach that utilizes a weighted average technique involving a set of n decision trees (DT). Furthermore, an intelligent feature selection algorithm that implements a recursive feature elimination method is employed. The optimal weights are determined using the Cuckoo Search algorithm. By combining the predicted model probability with its associated weight, the ensemble learning model achieves enhanced overall predictions. The evaluation process involves multiple ensemble learning-based decision trees, including Random Forest (RF), AdaBoost (AD), Gradient Boosting Tree Classifier (GB), and Extra Trees (ET). A credit card transaction dataset is randomly undersampled to evaluate the performance of the proposed solution. Through extensive experiments conducted on real data, the results highlight the superior predictive capabilities of the ensemble learning approach, demonstrating improvements in accuracy (ACC), precision (PRE), recall score (REC), and F-measure (F).
Keywords: Credit card, Ensemble learning, Cuckoo Search, Fraud detection, Dimensionality reduction.
Scope of the Article: Artificial Intelligence and Machine Learning