Visible to the public K-Means Clustering Analysis Based on Adaptive Weights for Malicious Code Detection

TitleK-Means Clustering Analysis Based on Adaptive Weights for Malicious Code Detection
Publication TypeConference Paper
Year of Publication2019
AuthorsHaoliang, Sun, Dawei, Wang, Ying, Zhang
Conference Name2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN)
PublisherIEEE
ISBN Number978-1-7281-2184-0
Keywordsadaptive weights-MMKM, AW-MMKM, clustering, composability, computer network security, cyber physical systems, k-means clustering analysis, malicious code, malicious code detection, malicious codes, Metrics, network behavior, network coding, network scanning, Network security, network traffic, pattern clustering, pubcrawl, resilience, Resiliency, statistical analysis, telecommunication traffic, traditional detection techniques, traffic characteristics
Abstract

Nowadays, a major challenge to network security is malicious codes. However, manual extraction of features is one of the characteristics of traditional detection techniques, which is inefficient. On the other hand, the features of the content and behavior of the malicious codes are easy to change, resulting in more inefficiency of the traditional techniques. In this paper, a K-Means Clustering Analysis is proposed based on Adaptive Weights (AW-MMKM). Identifying malicious codes in the proposed method is based on four types of network behavior that can be extracted from network traffic, including active, fault, network scanning, and page behaviors. The experimental results indicate that the AW-MMKM can detect malicious codes efficiently with higher accuracy.

URLhttps://ieeexplore.ieee.org/document/8905286
DOI10.1109/ICCSN.2019.8905286
Citation Keyhaoliang_k-means_2019