Comparative Analysis of Predicting Malware Attack Trends in Cyber Supply Chain Using Multiple Classification Models
S Li · IEEE Access · 2024
This paper investigates the prediction of malware attack trends within the cyber supply chain domain using data sourced from Microsoft Malware Prediction and Research Prediction for the year 2019. Various machine learning models, including XGBoost, Support Vector Classification (SVC), Extra Tree, K-Nearest Neighbors (KNN), Gradient Boosting, and a four-layer Neural Network, were trained and evaluated based on accuracy, precision, recall, and f1-score metrics. The findings demonstrate the Neural Network model performance of, surpassing other models in accuracy, recall, and f1-score, and closely trailing the Extra Tree technique in precision. Consequently, the Neural Network was identified as the most effective model for predicting malware attack trends within the cyber supply chain. This study underscores the significance of employing advanced neural network architectures in addressing complex cybersecurity challenges. The findings enhance cyber defense mechanisms and fortify critical infrastructure against evolving digital threats in supply chain networks.