AIpowered Automation of Fraud Detection in Financial Services
Rahul Sharma, Dattatraya Babanrao Nalawade, Pushpa Negi, Ritika Dhabliya, Saurabh Bhattacharya, Vinit Khetani · 2023
Due to the rise in credit card fraud, robust and efficient fraud detection systems are needed. This study introduces an AI-based banking fraud detection method. This paper introduces a new method that combines Support Vector Machines (SVM) with Random Forest algorithms to improve reliability. Credit card fraud detection is vital to protecting financial institutions and consumers. The study uses data preprocessing, feature engineering, and model selection to find the best fraud detection solution. Through wide experimentation and evaluation using real-world credit card transaction data, proposed hybrid model of SVM and Random Forest model achieved 99.42% accuracy. This performance shows the power of AI-driven automation and is a major improvement over current fraud detection method. Proposed method is highly precise and adaptable to new fraud patterns, making it ideal for the ever-changing financial services industry. Its real-time functionality makes it ideal for the dynamic industry. The hybrid model combines SVM and Random Forest benefits. This combination uses SVM classification accuracy and Random Forest's ensemble learning to enhance the model's overfitting resistance. AI-based solutions for financial services fraud detection have advanced significantly with this study. These solutions aim to increase security, reduce financial losses, and make the financial landscape safer for everyone. This enhances security and operational efficiency in digital finance's dynamic world.