Published Paper Details:

People rely nearly entirely on internet transactions in today's environment. While there are benefits to online transactions, such as ease of use, practicality, speedier payments, etc., there are drawbacks as well, such as fraud, phishing, data loss, etc. An individual's privacy may be violated by fraud and deceptive transactions, which are a continual concern with the rise in online transactions. In order to stop high risk transactions, several commercial banks and insurance providers invested millionsof rupees in developing a transaction detection system. We introduced a transaction fraud detection model with some feature engineering that is based on machine learning. As the algorithm processes as much data as it can, it will gain more stable and performance.The online fraud transaction detection project can make use of these methods. The se include the collection of a dataset including specific online transactions. Then, with the use of machine learning algorithms, we are able to identify the distinct or unusual data patterns that will be helpful in identifying fraudulent transactions. The KNN method, which consists of a cluster of decision trees, will be applied for optimal outcomes. Recently, this algorithm has taken over the ML community. This method is faster and more accurate than previous machine learning techniques.