Online Fraud Detection Using Machine Learning Algorithms and Quantum Annealing Solvers
B. Akshaya, Agnimita Das Bakshi, A. Annie Micheal · Advances in wireless technologies and telecommunication book series · 2024
Machine learning applications are widely used in identifying fraudulent transactions. Still, the majority of the applications detect fraudulent activities only after their occurrence, rather than in real-time or near real- time. The significant rarity of fraudulent transactions compared to normal ones creates a highly imbalanced data, posing substantial challenges for effective fraud detection. Addressing this issue extends beyond traditional machine learning approaches and requires alternative strategies. This chapter discusses the importance of fraud detection methods and compares k-nearest neighbor (KNN), random forest approach, artificial neural network (ANN) and support vector machine (SVM) enhanced with quantum annealing solvers that are used to detect fraud in online banking transactions. This study focuses on choosing the optimal approach for various types of datasets, while considering the three key factors such as accuracy, speed, and cost.