AI Enabled Credit Card Fraud Detection System Using Cloud Computing

Abstract

With the exponential growth of digital monetary transactions, the risk of credit card fraud has considerably extended, necessitating the improvement of advanced, real-time fraud detection structures. This studies aimed to design and compare a cloud-based answer powered by using synthetic intelligence (AI) to discover fraudulent transactions with high accuracy and low fake advantageous costs.A hybrid technique combining supervised system getting to know fashions and privacy engineering strategies was adopted to deal with demanding situations which include data imbalance and evolving fraud patterns. The method concerned the deployment of algorithms which include Support Vector Machines (SVM) and Naive Bayes (NB) on a synthetic yet sensible transaction dataset. Hybrid sampling strategies, which include a mixture of Synthetic Minority Oversampling Technique (SMOTE) and random underneath-sampling, were carried out to balance the dataset and improve model robustness.The gadget architecture was designed on a cloud platform to permit scalability and real-time transaction tracking. High-performance models have been deployed as APIs to method streaming statistics and provide instant class of transactions. Additionally, an Explainable AI (XAI) module was incorporated to enhance model transparency and guide regulatory compliance through supplying interpretable predictions.The findings confirmed that AI models should locate fraudulent interest in actual-time with over 99% accuracy, while additionally notably lowering fake positives. The use of cloud infrastructure facilitated dynamic resource scaling and ensured non-stop availability of the detection provider. The proposed framework proved effective in addressing key issues inclusive of integration complexity, information privateness, and real-time responsiveness.This have a look at concluded that the integration of AI, cloud computing, and explainable frameworks presents a scalable and sincere answer for modern-day credit score card fraud detection. Future directions include incorporating federated learning and blockchain technology to in addition decorate safety and transparency.