Comparative Study of k-Nearest Neighbor (KNN) and k-Means Algorithm in Fraud Detection

Abstract

Fraud detection especially in credit card is one of the challenging issues in these days. Finding irregularities is even more difficult due to high volume of data during the transaction. Many data mining techniques are applied by researchers for solving these problems. In this research, we explore K-Nearest Neighbor (KNN) and k-means algorithms which are widely used classification and clustering algorithms respectively. These algorithms are used in this research to find out the better among them. Moreover, we also optimize the system by finding the most dominant and influencing factor responsible for fraud which will help in effective fraud detection in case of credit card.