Applications of Machine Learning in Fintech Credit Card Fraud Detection
Francisco Agramunt Lacruz, Jafar Saniie
Abstract
Fintech utilizes innovative technology to offer improved monetary administrations and financial solutions. According to data from the prediction of Autonomous Research artificial intelligence (AI) technologies will allow financial institutions to reduce their operational costs by 22% by 2030. Throughout this paper, we study how AI and machine learning algorithms can lead to credit card fraud detection. After making the theoretical approach to the subject, we develop two different methods Autoencoder (semi-supervised learning) and Logistic Regression (supervised learning) for fraud detection with a high level of accuracy. The results obtained with both methods are promising as we were able to predict fraud transactions with 94% certainty.