Published Paper Details:

Scam using a credit card is a simple and convenient case. E-commerce and many other paid internet sites have increased their online modes of operation, which increases the risk of online fraud. As the rate of fraud has increased, researchers have begun to use a variety of techniques used to train fraud detection and analysis in online transactions. The main goal is to project, develop and develop a new fraud detection method Streaming data about operations that are aimed at detailed analysis of the most recent customer operations and extracting data about people's behavior. This work was carried out by a dataset of credit card transactions entering the EU, a card containing 284,807 operations. In this project, we used an algorithm in an isolated forest for the purpose of classifying an observation. As a result, the model allows sample training to classify test select. Therefore, once the model is trained, we had to draw in the accuracy of the results obtained