Advanced Machine Learning Techniques Revolutionize Credit Card Fraud Detection

The research investigates the use of new machine learning techniques, including Decision Trees, Random Forests, and Extreme Gradient Boosting, to detect and prevent credit card fraud. The strength and performance of the models are compared through their validation using publicly available data as we...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:2025 International Conference on Networks & Advances in Computational Technologies (NetACT) s. 1 - 6
Hlavní autori: A, Balamanikandan, D, Jayakumar, M, Dhanalakshmi, G, Sudharsan, Prabha, Ankala Satya, V, Venkata Karthika
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 07.08.2025
Predmet:
Accuracy
Boosting
CNN
Convolutional neural networks
Credit cards
Data models
Decision Tree
Decision trees
Deep learning
Extreme Gradient Boosting Algorithms
Feature extraction
Fraud
MLP
Random Forest
Random forests
stacking and voting classifier
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:The research investigates the use of new machine learning techniques, including Decision Trees, Random Forests, and Extreme Gradient Boosting, to detect and prevent credit card fraud. The strength and performance of the models are compared through their validation using publicly available data as well as real credit card data provided by financial institutions. The robustness of the system is verified by adding noise to the data samples. Major findings indicate that such machinelearning algorithms are very effective in identifying fraud, with the CNN model showing to have an accuracy rate of almost 99 %. The report indicates the necessity for sophisticated fraud detection systems given heightened online payments and evolving patterns of fraud. The research finds that the deep learning techniques, particularly CNNs, are highly effective for the detection of credit card fraud, identifying fine patterns in transactional data, and enhancing the accuracy of predictions. The research highlights the importance of real-time monitoring and updating of fraud detection systems so that they can learn to adapt to new developing fraud patterns, making the systems effective and reliable in real-world applications.
DOI:10.1109/NetACT65906.2025.11188095