Credit Card Transaction Validity Detection Using Density Based Clustering

Abstract

The use of credit cards is increasing every year, causing abuse of credit card transactions is also increasing. The perpetrators of crime with credit card fraud are also increasingly sophisticated and have a wide network. This causes a loss for all parties. To detect credit card misuse in electronic transactions, the author tries to make research using Density Based Clustering approach (DBSCAN) to detect whether credit card number is valid or not. DBSCAN generally starts by setting the starting point randomly, then finding all the dots around in Eps the starting point distance, if the number of dots around is greater than or equal to MinPts then the cluster is formed, using density distributions from the dots in the database, DBSCAN can categorize the points into separate groups indicating different classes. It is hoped that by using DBSCAN Method in the decision making process can know the existence of misuse in the use of transaction with credit card. DBSCAN method in this research is used to help determine credit card problem in electronic transaction by using parameter of minimum point (minpts) and epsilon (eps). So it can be concluded whether the credit card is valid or not.