.png){kind=icon}
Return
Performance Analysis of Visualization-Based Malware Classification Using CNN
.png){kind=icon}
PRE
AI
DOI：10.1080/03772063.2025.2561713.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With the rapid expansion of the cyber world, incidents of cyberattacks and malware threats such as viruses, worms, trojans, and ransomware have increased significantly. Traditional signature-based intrusion detection system struggles to detect novel and unknown malware variants. Anomaly-based network intrusion detection system often fails against adversarial attacks involving fragmented and encrypted malicious data packets. A deep-learning-based one-dimensional CNN (1D-CNN) is employed to detect and classify adversarial variants of novel malware using visualization techniques. This study evaluates the effectiveness of 1D-CNN, transfer learning, generative adversarial networks (GANs), and ensemble learning techniques on the Malimg dataset and the IEEEDataPort binary-class dataset. Deep-learning models are assessed based on classification accuracy, training time, and computational cost. The models include custom and classical 1D-CNN architectures, enhanced EfficientNet-based transfer learning models, conditional GANs, and stacked ensemble techniques, achieving a classification accuracy of up to 99.25%. Performance comparisons with state-of-the-art research demonstrate that the proposed methods outperform many existing techniques in efficiency, accuracy, and computational cost.
Keywords:
CNN
GAN
Information security
Malware classification
transfer learning
Journal
IF:
1.3
Papers:
261
Citations:
3.3K
Organization
Cited Papers
A Lightweight Generative Adversarial Network for Imbalanced Malware Image Classification.png)
Proceedings of the 5th International Conference on Information Management & Machine Intelligence
OAAI
Proceedings of the 5th International Conference on Information Management & Machine IntelligenceIF0
OAAIShareSaveShareSaveEnhanced Image-Based Malware Classification Using Snake Optimization Algorithm With Deep Convolutional Neural NetworkIEEE ACCESSOAAI
IEEE ACCESSIF3.6
OAAIShareSaveShareSave
ShareSaveShareSaveShareSave
View more