Fraud detection in credit card transaction using hybrid model
A. Maria Nancy, G. Senthil Kumar, S Veena, N. A. S. Vinoth, Moinak Bandyopadhyay
Abstract
Credit card payment is a popular mode of online transaction. It is one of the simplest and easiest mode of payment across the internet. However, with growing popularity of credit card transactions, there is an exponential growth in fraudulent payments. Every year we lose billions of dollars due to fraudulent acts. These activities look like a genuine transaction; hence, simple pattern techniques and less complex methods don’t notgo to work to minimize the fraudulent act and minimize the chaos we are proposing a Hybrid model consisting of Deep Learning Algorithm of Convolutional Neural Network followed by K- Nearest Neighbors Classification. These two approaches are proved to decrease the false alarm rates and increase the fraud detection rate and expected to be more efficient than other relevant algorithms.