Fake Transaction Detection in Credit Card using Machine Learning
The active participation coupled with warp speed engagement in online-based transactional activities raise the emergence of fake cases all over the world. This causes tremendous losses to the individuals as well as banking industry. Although there are many criminal activities occurring in financial sector, credit card activities of fake nature are among the most prevalent causing huge worryto online customers. Countering these nefarious actions through Data Mining (DM)and Machine Learning(ML)techniques is one of the prominent approaches introduced by scholars intending to prevent the damage caused by these illegal acts. Primarily, DM techniques are employed to study the patterns and characteristics of suspicious vis-à-visnon-suspicious transactions based on normalized and anomalies data. On the other hand, ML techniques are employed to predict automatically the suspicious and non-suspicious transactionsby using classifiers. A rich combination of DM and ML techniques are able to identify the genuine and spurious transactions by learning the patterns of the data. Thisdiagnostic Paper discusses the supervised-based classification using Binary Classification namely Regression and Random Forest. After prior-processing the dataset by employing normalization and Principal Component Analysis, all the classifiers achieved more than 99.90% accuracy compared to the results attained before pre-processing the dataset.
Keywords - Credit Card; Fake Detection; Machine Learning.