Quantum Exploration in Ransomware Detection with Conventional Machine Learning Approaches
Manvi Kaur, Kriti Jain, Ananya Singla, Karuna Kadian · 2024
Ransomware attacks are a significant and evolving cybersecurity threat that requires innovative and efficient detection methods. This study provides a comprehensive analysis of the performance of quantum machine learning (QML) models versus traditional machine learning techniques. In particular, the focus is on the evaluation of three QML models (VQC, Pegasos QSVM, and Quantum-augmented SVM) that have not yet been studied in the context of ransomware detection. This study examines in detail the effectiveness and resource utilization of traditional machine learning models (SVM, decision trees, XGBoost) using a dynamic ransomware database. To assess the efficacy of our model, we compare performance metrics like F1 score, recall, accuracy, and precision. The study also illustrates how training and testing times for quantum and classical models differ. The study addresses current problems and highlights the potential and limitations of OML.