Malware Image Classification using Machine Learning with Local Binary Pattern

<p>Malware classification is a critical part in the cybersecurity.</p> <p>Traditional methodologies for the malware classification</p> <p>typically use static analysis and dynamic analysis to identify malware.</p> <p>In this paper, a malware classification methodology based</p> <p>on its binary image and extracting local binary pattern (LBP)</p> <p>features are proposed. First, malware images are reorganized into</p> <p>3 by 3 grids which is mainly used to extract LBP feature. Second,</p> <p>the LBP is implemented on the malware images to extract features</p> <p>in that it is useful in pattern or texture classification. Finally,</p> <p>Tensorflow, a library for machine learning, is applied to classify</p> <p>malware images with the LBP feature. Performance comparison</p> <p>results among different classifiers with different image descriptors</p> <p>such as GIST, a spatial envelope, and the LBP demonstrate that</p> <p>our proposed approach outperforms others.</p>

Name thesis:degree_name

Master of Science in Computer Science (MSCS)

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Year dc:date.available

2018

Authors dc:creator

• Luo, Jhu-Sin
• Lo, Dan

Contributors dc:contributor

• Dan Lo

• malware
• classification
• machine learning
• visualization
• local binary pattern
• Information Security

Repository record dc:identifier

https://digitalcommons.kennesaw.edu/cs_etd/16

OAI identifier oai:identifier

oai:digitalcommons.kennesaw.edu:cs_etd-1019