IRJET- Credit Card Fraud Detection Techniques (original) (raw)

Abstract

Credit card frauds are easy and friendly targets. E-commerce and many other online sites have increased the online payment modes, increasing the risk for online frauds. Increase in fraud rates, researchers started using different machine learning methods to detect and analyse frauds in online transactions. Frauds are known to be dynamic and have no patterns, hence they are not easy to identify. Fraudsters use recent technological advancements to their advantage. They somehow bypass security checks, leading to the loss of millions of dollars. Analyzing and detecting unusual activities using data mining techniques is one way of tracing fraudulent transactions. transactions. This paper gives a survey of multiple machine learning methods such as k-nearest neighbor (KNN), random forest, naive bayes, logistic regression and support vector machines (SVM) as well as the deep learning methods such as autoencoders, convolutional neural networks (CNN), restricted boltzmann machine (RBM) and deep belief networks (DBN).

Key takeaways

AI

1. Machine learning methods effectively address dynamic credit card fraud detection challenges, especially under class imbalance.
2. This paper surveys various techniques including KNN, Random Forest, and deep learning models like CNN and Autoencoders.
3. Supervised learning algorithms yield promising results but struggle with evolving fraud patterns over time.
4. Meta-classification strategies can improve performance by combining predictions from multiple classifiers, achieving a 28% improvement.
5. Larger datasets favor SVMs and CNNs for reliable fraud detection outcomes, as shown in empirical comparisons.

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (9)

1. Credit Card Fraud Detection using Machine Learning and Data Science by S P Maniraj , Aditya Saini, Swarna Deep Sarkar and Shadab Ahmed, International Journal of Engineering Research & Technology (IJERT), Vol. 8 Issue 09, September- 2019. Available: www.ijert.org
2. Credit card fraud detection using Machine learning algorithms by AndhavarapuBhanusri, K.RatnaSree Valli , P.Jyothi , G.Varun Sai , R.Rohith Sai Subash, Journal of Research in Humanities and Social Science Volume 8 ~ Issue 2 (2020). Available:www.questjournals.org
3. Credit Card Fraud Detection using Machine Learning Algorithms by Vaishnavi Nath Dornadulaa and Geetha S, international conference on recent trends in advanced computing 2019, icrtac 2019. Available: www.sciencedirect.com
4. Fraud Detection using Machine Learning and Deep Learning by Pradheepan Raghavan and Neamat El Gayar, 2019 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE) December 11-12, 2019, Amity University Dubai,UAE. Available: ieeexplore.ieee.org
5. Credit card fraud detection using Machine Learning Techniques by John O. Awoyemi, Adebayo O. Adetunmbi and Samuel A. Oluwadare
6. Tom Sweers. "Auto encoding Credit Card Fraud". Bachelor Thesis, Radboud University, June 2018.
7. Alae Chouiekha, EL Hassane Ibn EL Haj. "ConvNets for Fraud Detection analysis". Procedia Computer Science 127, pp.133-138. 2018.
8. S. Maes, K. Tuyls, B. Vanschoenwinkel, B. Manderick. "Credit Card Fraud Detection Using Bayesian and Neural Networks". 2002. [Online]. Available: https://www.researchgate.net/publication/252470 7_Credit_Card_Fraud_Detection_Using_Bayesian_and _Neural_Networks.
9. Stolfo, S., Fan, D. W., Lee, W., Prodromidis, A., & Chan, P. (1997). Credit card fraud detection using meta-learning: Issues and initial results. In AAAI-97 Workshop on Fraud Detection and Risk Management

FAQs

AI

What explains the class imbalance in credit card fraud detection datasets?add

The paper notes that valid transactions vastly outnumber fraudulent ones, creating challenges for detection algorithms. For instance, fraudulent transactions are often rare, making supervised learning less effective without modifications.

How do different algorithms compare in detecting credit card fraud?add

The study empirically compares the performance of various algorithms, finding that SVMs and CNNs perform best with larger datasets. Notably, a meta-classifier approach showed a 28% improvement over baseline models.

What challenges arise with using deep learning for fraud detection?add

The paper indicates deep learning models, like Autoencoders, struggle with smaller datasets and yield poorer prediction scores. Additionally, such models may fail to adapt as fraud patterns evolve over time.

When should supervised learning methods be applied to fraud detection?add

Supervised learning methods are most suitable for static datasets where fraud patterns are stable. Despite good performance, methods like KNN and Random Forest are less effective in dynamic environments.

What recent advancements have improved fraud detection strategies?add

The implementation of hybrid models combining Adaboost and Majority Voting has demonstrated enhanced detection accuracy. Further, innovations like Artificial Genetic Algorithms have improved the minimization of false alerts.