Fraud Detection in Online Payments using Machine Learning Techniques
U. Siddaiah, P. Anjaneyulu, Yanamala Haritha, M Ramesh · 2023
People rely on online transactions for nearly everything in today’s environment. Online transactions offer several benefits, such as ease of use, viability, speedier payments, etc., but they also have some drawbacks, such as fraud, phishing, data loss, etc. As online transactions grow, there is a continuing risk of frauds and deceptive transactions that could violate a person’s privacy. In order to prevent high risk transactions, numerous commercial banks and insurance firms invested millions of rupees in the development of transaction detection systems. This research study has introduced a feature-engineered machine learning-based model for detecting transaction fraud. By processing as much data as it can, the algorithm can gain experience, strengthen its stability, and increase its performance. The effort to detect online fraud transactions can use these algorithms. In this, a dataset of specific online transactions is obtained. Then, with the aid of machine learning algorithms, unusual or distinctive data patterns that will be helpful in identifying any transactions that are fraudulent are discovered. The XGBoost algorithm is a cluster of decision trees, which will be utilized in order to achieve the best outcomes. This algorithm has recently taken control of the ML industry. Comparing this approach to other ML algorithms reveals that it is faster and more accurate.