Malware detection using linear SVM
Authors:
- 1 June 2013
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 136-138
Abstract
The governments of developed countries already have the policy framework of national anti-virus software addressing the security issues and developing nations also tend to follow this trend - making comprehensive effort on anti-virus software development.For our country, we are facing with the challenge to develop this “strategic technology” and create anti-virus software framework and resources in a next few years, which is one of the national security wide concerns. The detection of malware is the most significant part of malware protection. In this paper, we provide a “data mining” approach for malicious software detection and performed some experimental investigation on malware detection using linear SVM algorithm. The goal of this work is to show actual result of malware detection rates of SVM method.The SVM classifier is approved to detect unknown samples of malware with the probability of 74 - 83 percent. The detection principle is that, SVM algorithm generates detection model learning from the sufficient dataset of malicious software.Keywords
This publication has 4 references indexed in Scilit: