Analysis of Fraud Detection in Online Transactions Using Computational Models

Financial fraud is becoming a more serious menace to the financial industry. One important tool for identifying credit card fraud in online payments is data mining. There are two primary reasons why credit card fraud, a data mining issue, becomes challenging to identify: First of all, the characteristics of typical conduct and fraud are ever-evolving. Second, credit card data sets are highly corrupted, to start with. Theft or misuse of your credit card information for personal use without your knowledge or agreement is known as credit card fraud. User behavior in previous transactions must be verified to detect these frauds. It can be classified as a fraud or a legitimate transaction when comparing its usage pattern with that of existing transactions. The variables chosen, the data set measurement strategy, and the detection techniques employed all have a major impact on the capacity to identify fraud in credit card purchases. It seems that information extraction has become the main task in detecting online payment fraud. For highly distorted credit card fraud data, this article uses K-Nearest Neighbor and Logistic Regression techniques.