Parallelisation of machine learning models for credit card fraud detection Online publication date: Mon, 23-Oct-2023
by J. Saira Banu; Sumaiya Thaseen Ikram; Vanitha Mohanraj; Kishore Sanapala; Polinpapilinho F. Katina
International Journal of System of Systems Engineering (IJSSE), Vol. 13, No. 4, 2023
Abstract: The use of credit cards for financial transactions is widely spreading in the current era. Credit card usage and fraud transactions are proportional to each other. This research analyses the impact of machine learning techniques such as base classifiers and ensemble models to detect fraud transactions. This is considered as a system of systems as there are a collection of many independent base and ensemble models that work together for prediction. Hyper-parameter tuning is an integral activity for enhancing model prediction. Two types of parameter tuning are performed: random and grid search. In addition, a comparative analysis is performed to evaluate the impact of cores. The Kaggle credit card fraud identification dataset is utilised for experimental analysis. While the analysis indicates that the training time of 5-classifiers is diminished, it has not been parallelised with 12-classifiers. The experimental analysis has an accuracy of 99.9% for 5-classifiers, including random and ensemble classifiers with hyperparameter tuning. The system's performance is also estimated using derived metrics such as precision, recall, f-score, and Matthews correlation coefficient.
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of System of Systems Engineering (IJSSE):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email subs@inderscience.com