MENDETEKSI KECURANGAN PADA TRANSAKSI KARTU KREDIT UNTUK VERIFIKASI TRANSAKSI MENGGUNAKAN METODE SVM

Abstract

Credit cards are one of the most popular and many used payment methods on online transactions. In line with the large number of credit card users and even become a daily payment method, the security in verifying every transaction is also very important to be improved. Data mining is one of the most method that can assist in solving problems in credit card transaction security. This research purpose to design a verification system of credit card transactions, which can help the banks in knowing the possibility of fraud that occurred. This research uses support vector machine (SVM) method to detect fraud based on outlier / anomaly on transaction data.Sample data of 100 rows, using attribute of account number as label, month, and transaction nominal as training data. Testing is done by simulating a transaction on the form, by filling in the account number, month, and transaction nominal. The system will detect whether tranksaction with the account number is entered in the designated in the model class. If out of the class then the transaction is suspended.