Fraud recognition in Digital Transactions by using SMOTE algorithm
Full Text |
Pdf
|
Author |
V. Priyadarshini and A. Pushpa Latha
|
e-ISSN |
1819-6608 |
On Pages
|
552-562
|
Volume No. |
18
|
Issue No. |
05
|
Issue Date |
April 05, 2023
|
DOI |
https://doi.org/10.59018/032379
|
Keywords |
fraud detection (FD), digital transactions, machine learning (ML), data mining (DM), credit card fraud (CCF), synthetic minority oversampling approach (SMOTE).
|
Abstract
Digital transactions make our lives easier. At the same time, every human is facing fraud issues by using Digital Transactions like credit cards with the growing number of transactions. Many intruders try to steal credit card details using various internet sources and cheat credit card holders. The intruders play unique tricks to cheat users, like sending trustworthy messages and emails. An enhanced fraud detection technique has become necessary to keep users sustainable to overcome the problem. An Ensemble model is constructed in this study utilizing the SMOTE algorithm to detect fraudulent transactions and alert users. The model performance is evaluated by using Machine Learning Models like KNN, Logistic Regression, and SMOTE. Among these, the SMOTE algorithm has the highest accuracy in detecting fraud.
Back