Publisher

IDID

Journal Jurnal Ilmu Komputer dan InformatikaJurnal Ilmu Komputer dan Informatika Abstract

Credit card fraud is a growing problem due to the rise of card transactions. This study investigates the effectiveness of Logistic Regression (LogReg) and Extreme Gradient Boosting (XGBoost) in identifying fraudulent transactions in a highly imbalanced dataset, where only 8% of the data represents fraudulent activity. To address the class imbalance, random undersampling was applied, reducing the number of legitimate transactions. This technique significantly improved LogRegs ability to detect fraud, with the AUC-ROC increasing from 0.7994 to 0.9089. XGBoost performed well even without hyperparameter tuning or random undersampling, indicating its robustness as a baseline model. The study highlights the critical importance of addressing class imbalance in fraud detection. Both LogReg and XGBoost demonstrated potential, particularly when combined with techniques like undersampling or hyperparameter tuning. These findings underscore the need for effective data preprocessing methods to enhance the performance of machine learning models in detecting credit card fraud.

Conclusion

This study demonstrates the importance of addressing class imbalance in credit card fraud detection.Logistic Regression benefits significantly from random undersampling, improving its ability to identify fraudulent transactions.XGBoost exhibits robustness and strong performance even without extensive tuning, making it a reliable baseline model.These findings emphasize the need for effective data preprocessing techniques to optimize machine learning models for fraud detection.

Future Research

Further research could explore the combination of random undersampling with other data balancing techniques, such as SMOTE, to potentially enhance model performance further. Investigating the application of more advanced hyperparameter optimization algorithms, like Bayesian optimization, could lead to even more refined XGBoost models. Additionally, a study focusing on the interpretability of XGBoost models, using techniques like SHAP values, could provide valuable insights into the key features driving fraud detection and improve trust in the models predictions. These investigations will contribute to the development of more effective and reliable fraud detection systems.

References

1. Unbalanced Data Processing and Machine Learning in Credit Card Fraud Detection | Research Square. unbalanced... researchsquare.com/article/rs-2004320/v1Unbalanced Data Processing and Machine Learning in Credit Card Fraud Detection Research Square unbalanced researchsquare article rs 2004320 v1
2. Customer Transaction Fraud Detection Using Xgboost Model | IEEE Conference Publication | IEEE Xplore.... ieeexplore.ieee.org/document/9103880Customer Transaction Fraud Detection Using Xgboost Model IEEE Conference Publication IEEE Xplore ieeexplore ieee document 9103880
3. SciELO Brazil - Read this paper if you want to learn logistic regression Read this paper if you want... doi.org/10.1590/1678-987320287406ENSciELO Brazil Read this paper if you want to learn logistic regression Read this paper if you want doi 10 1590 1678 987320287406EN

Metadata

File size	455.69 KB
Pages	12
DMCA	Report