This summary outlines the investigation into detecting and categorizing cloud computing malware using Artificial Neural Network (ANN)-based Machine Learning models. The surge in cloud services has elevated the risk of malware attacks in cloud environments. This research explores the application of ANN-based ML techniques to bolster cloud system security, emphasizing robust malware detection through feature extraction, dataset preparation, and model training. Various ANN architectures are considered and evaluated against real-world cloud malware datasets, revealing the efficacy of ANN models in detecting and classifying cloud-based malware with promising accuracy and efficiency. The deployment of a neural network classifier for binary malware that classifies Windows Portable Executable (PE) files according to imported library function calls is the specific topic of this article. The applied model demonstrates its capacity to generalize against an independent set with an astounding 97.8% average accuracy, 97.6% precision, and 96.6% recall. These findings demonstrate the practicality of the suggested malware categorization technique for bolstering cloud security against changing cyber threats.
Paper
Full text
Dual attention-based cloud virtualization security using machine learning and big data
Semantic Scholar · Computer Science · 2025
Abstract
This summary outlines the investigation into detecting and categorizing cloud computing malware using Artificial Neural Network (ANN)-based Machine Learning models. The surge in cloud services has elevated the risk of malware attacks in cloud environments. This research explores the application of ANN-based ML techniques to bolster cloud system security, emphasizing robust malware detection through feature extraction, dataset preparation, and model training. Various ANN architectures are considered and evaluated against real-world cloud malware datasets, revealing the efficacy of ANN models in detecting and classifying cloud-based malware with promising accuracy and efficiency. The deployment of a neural network classifier for binary malware that classifies Windows Portable Executable (PE) files according to imported library function calls is the specific topic of this article. The applied model demonstrates its capacity to generalize against an independent set with an astounding 97.8% average accuracy, 97.6% precision, and 96.6% recall. These findings demonstrate the practicality of the suggested malware categorization technique for bolstering cloud security against changing cyber threats.