Fraud Detection of Credit Card using Data Mining Techniques

A credit card is an essential and daily used component in our life. Credit card fraud is a crime in our society. It causes many problems in our daily online transactions. Various deficiency of the previous method has been researched in the introduction chapter. The establishment of avoiding credit card fraud security is a big issue for small financial institutions. The goal of this paper is to propose a model for detecting credit card fraud and making a false alarm. For this research purpose, we have used four data mining techniques i.e. Decision Tree (DT), Random Forest (RF), Artificial Neural Networks (ANN), and Logistic regression (LR). Our motivation was how to reduce the false alarm so that the cardholder gets into less trouble and also the card provider gets more time for accurate alarming checking. After collecting the data from the Kaggle repository we look into the structure of the large dataset to check the performance. Datasets are filtered by the feature selection algorithms called Pearson correlation and chi-squared. We have introduced this model to get fewer false alarms in credit card fraud. From the proposed system, we can obtain about 99% accuracy (LR) and it also gives fewer false alarms. For tracking credit card fraud in the era of industry 4.0, our research work will ascertain the advancement and effectiveness of sustainable technologies.