Fraud Detection in Credit Card Transactions Using SVM and Random Forest Algorithms

• Published in: 2021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC) pp. 1013 - 1017
• Main Authors: Saddam Hussain, S K, Sai Charan Reddy, E, Akshay, K Gangadhar, Akanksha, T
• Format: Conference Proceeding
• Language: English
• Published: IEEE 11.11.2021
• Subjects: Costs; credit card; Credit cards; Data models; data set visualization; Data visualization; Decision Trees; fraud detection; Graphics; Sensitivity; Support vector machines
• ISSN: 2768-0673
• DOI: 10.1109/I-SMAC52330.2021.9640631

This project's primary objective is to detect credit card fraud in the real world. Recent growth has resulted in a significant increase in the number of credit card transactions. The objective is to obtain goods from an account without paying for them or using unapproved funds. It is critical for all banks that issue credit cards to reduce the cost of implementing an effective fraud detection system. One of the most difficult challenges is that neither the card nor the cardholder is needed to complete the transaction during a credit card transaction. Thus, the seller cannot verify whether the customer who is making an acquisition is an authentic cardholder or not. The accuracy of detecting fraud is improvised with this system proposed using random forest algorithm, decision tree, and support vector machine algorithms. A random forest algorithm is a classification process for observing the data set and optimizing the accuracy of the resultant data. The techniques' performance is judged based on precision, sensitivity, & accuracy. Some of the data provided are processed to identify fraud detection and provide visualization for the graphic model.