An Efficient Real Time Model For Credit Card Fraud Detection Based On Deep Learning

In the last decades Machine Learning achieved notable results in various areas of data processing and classification, which made the creation of real-time interactive and intelligent systems possible. The accuracy and precision of those systems depends not only on the correctness of the data, logically and chronologically, but also on the time the feed-backs are produced. This paper focuses on one of these systems which is a fraud detection system. In order to have a more accurate and precise fraud detection system, banks and financial institutions are investing more and more today in perfecting the algorithms and data analysis technologies used to identify and combat fraud. Therefore, many solutions and algorithms using machine learning have been proposed in literature to deal with this issue. However, comparison studies exploring Deep learning paradigms are scarce, and to our knowledge, the proposed works don't consider the importance of a Real-time approach for this type of problems. Thus, to cope with this problem we propose a live credit card fraud detection system based on a deep neural network technology. Our proposed model is based on an auto-encoder and it permits to classify, in real-time, credit card transactions as legitimate or fraudulent. To test the effectiveness of our model, four different binary classification models are used as a comparison. The Benchmark shows promising results for our proposed model than existing solutions in terms of accuracy, recall and precision.