Real Time Credit Card Fraud Detection and Reporting System Using Machine Learning

Simple item page
dc.contributor.authorAhmed Oluwatoyin JOLAOSHO
dc.date.accessioned2025-10-10T13:33:54Z
dc.date.available2025-10-10T13:33:54Z
dc.date.issued2023-12
dc.description.abstractIn recent years, there has been a significant rise in fraudulent credit card activities, resulting in substantial financial losses for numerous organizations, companies, and government agencies. This study addresses the critical issue of credit card fraud detection in real-time using machine learning algorithms. The primary objective is to develop a robust prototype model capable of promptly identifying fraudulent transactions and notifying users. To achieve this, three algorithms, namely the Random Forest, the Decision Tree classifier and Linear Regression algorithms were used. Furthermore, various sampling techniques were employed to balance the dataset and improve model performance. 12 distinct models were developed, each offering varying levels of accuracy and effectiveness in fraud detection. From the findings, transactions are most commonly made after 12 noon. Also, older individuals above 75 years are more susceptible to fraud, possibly due to their unfamiliarity with evolving transaction methods. Also, transactions in the dataset are predominantly made the Female gender suggesting that transactions involving this gender may be more prone to fraud. Findings also showed that among these models, the Random Forest -SMOTE [Hyperparameter Tuned], emerged as the best classifier with remarkable performance metrics, including a 97% accuracy rate, an F1 score of 95%, and a precision rate of 98%. The study extended its focus to practical implementation by integrating the Random Forest -SMOTE [Hyperparameter Tuned] with Twilio for real-time notification. This integration successfully demonstrated the model's ability to send timely and accurate fraud alerts to users. The analysis and model development for fraud detection has therefore provided valuable insights and a robust solution for real time identifying and responding to fraudulent activities. It is recommended that periodic evaluations of the fraud detection model's performance be performed to ensure its effectiveness in detecting evolving fraud patterns. Keywords: Accuracy, Credit Card Fraud Detection, Decision Tree Classifier, Fraud Detection Model, Fraudulent Transactions, Real-time Notification, Sampling Techniques Word Count: 289 Words
dc.identifier.citationKate Turabia
dc.identifier.otherM.Sc
dc.identifier.urihttps://repository.lcu.edu.ng/handle/123456789/1183
dc.language.isoen
dc.publisherLead City University, Ibadan
dc.relation.ispartofseriesM.Sc
dc.subjectAccuracy
dc.subjectCredit Card Fraud Detection
dc.subjectDecision Tree Classifier
dc.subjectFraud Detection Model
dc.subjectFraudulent Transactions
dc.subjectReal-time Notification
dc.subjectSampling Techniques
dc.titleReal Time Credit Card Fraud Detection and Reporting System Using Machine Learning
dc.typeThesis

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Jolaosho, Ahmed Oluwatoyin FULL THESIS.pdf
Size:
6.52 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed to upon submission
Description:
Download

Collections

Department of Computer Science;