Year: 2025
Type: article
Abstract: Financial fraud has become a growing concern in the digital era, with cybercriminals leveraging sophisticated techniques to exploit vulnerabilities in financial transactions. Traditional rule-based fraud detection methods often fail to adapt to evolving fraudulent patterns, necessitating advanced data-driven approaches. This research explores anomaly detection in financial transactions using machine learning and data analytics techniques. By analyzing transaction patterns, our study leverages unsupervised learning models such as Isolation Forest, Autoencoders, and One-Class SVM to identify deviations indicative of fraudulent activity. Additionally, supervised models like Random Forest and XGBoost are employed for comparative evaluation. The dataset consists of real-time financial transactions, and feature engineering techniques, including transaction frequency, amount variation, and location-based anomalies, are applied to enhance detection accuracy. The proposed methodology is tested on benchmark datasets, demonstrating its effectiveness in reducing false positives while improving precision and recall. The study also highlights the importance of explainability in AI-driven fraud detection systems to ensure transparency in decision-making. The findings suggest that integrating data analytics in financial security can significantly enhance fraud prevention mechanisms, providing banks and financial institutions with proactive strategies to mitigate financial risks.
Author: Shamna M
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4409173234