Multiple Classifier Systems for More Accurate JavaScript Malware Detection

DOI

10.2991/icpit-16.2016.22How to use a DOI?

Keywords

machine learning, JavaScript malware detection, multiple classifier system

Abstract

The researches of JavaScript malware detection focus on machine learning techniques in recent years. These works extract features from JavaScript's abstract syntax tree for the training of classifiers and achieve satisfactory detection results. However, in the training set there exist some scripts that are not so representative and may cause occasional incorrect classification. We propose multiple classifier system (MCS) to reduce this kind of misclassification. As shown in the experiments, the accuracy increases because of the MCS while training time is slightly greater than the original classifier.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Zibo Yi
AU  - Jun Ma
AU  - Lei Luo
AU  - Jie Yu
AU  - Qingbo Wu
PY  - 2016/08
DA  - 2016/08
TI  - Multiple Classifier Systems for More Accurate JavaScript Malware Detection
BT  - Proceedings of the International Conference on Promotion of Information Technology (ICPIT 2016)
PB  - Atlantis Press
SP  - 139
EP  - 143
SN  - 2352-538X
UR  - https://doi.org/10.2991/icpit-16.2016.22
DO  - 10.2991/icpit-16.2016.22
ID  - Yi2016/08
ER  -