The current digital technology is emphasized in every part of the society which enhanced ecommerce, m commerce, and online shopping market places. Some financial enterprises have introduced ease way of acquiring credit card to its customers with the credit score facility. Thus more number of financial transactions were been performed using card system or enhanced payment mechanisms. It triggered an increase in the credit card anomalous transfers. Such fraud need to be identified and controlled. Many methods based on detection were proposed towards the problem of eradicating frauds. Hereby, it is addressed to solve the said issue initially using classification of normal/fraud values with machine learning technique. To extract the features that distinguish fraud and genuine transactions, the data mining techniques are found to be applied appropriately. The implementation part is achieved with decision tree methodology in order to classify normal and fraud data. The metric considered is based on Gini index that is been evaluated during the splits of the decision tree. The proposed work is well observed with respect to the accuracy of the misclassified samples. In this paper, a classification model based is thus utilized to train the features and predict the fraudulent and genuine transactions.
Volume 12 | 08-Special Issue
Pages: 530-537
DOI: 10.5373/JARDCS/V12SP8/20202552