Year: 2025
Type: conference-paper
Abstract: This paper proposes a combined model for JavaScript malicious code detection based on deep learning Q - network to address the severe network security situation where JavaScript malicious codes are rampant. The model fully exploits the advantages of deep learning Q - network in processing sequential data and complex decision - making tasks. Through learning a large number of JavaScript code samples, an intelligent detection system capable of accurately identifying malicious codes is constructed. The model innovatively combines multiple detection techniques, including static feature extraction and dynamic behavior analysis, to comprehensively capture the characteristic information of malicious codes. Experimental results show that, compared with traditional detection methods, this combined model significantly improves the accuracy of JavaScript malicious code detection, effectively reduces the false positive rate and false negative rate, and demonstrates good robustness and adaptability. This research provides a new and effective solution for malicious code detection in the field of network security, contributing to ensuring the security and stability of the network environment.
Authors: Wenming Wang, Hongru Song, Jiaolong Zheng
Language: English
Subfield: Signal Processing
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4414288586