Fraud Classification and Detection Model Using Different Machine Learning Algorithm
Main Article Content
Abstract
Recently, fraud technologies have become more advanced and easier to fraud. Therefore, different machine learning techniques have been applied and developed to recognize fraudulent credit card transactions. The main problem to fail any detection techniques on any fraud operation is the accuracy of results. This paper discusses how to improve fraud detection performance using machine learning algorithms by choosing the most appropriate algorithm for inclusion in fraud detection systems. It also provides a comprehensive study of Taiwan's customer database and how classifiers interact with it by applying 30 different classification algorithms. Moreover, using the WEKA tool for applying machine learning algorithms with the voting method to choose the right classification. The experimental results reveal that using the LMT algorithm will be the best one where achieved 82.0867 % accuracy.