Android malicious code detection and recognition based on depth learning

DOI

10.2991/macmc-17.2018.40How to use a DOI?

Keywords

malicious code; detection algorithm; depth learning; neural network; Andriod terminal

Abstract

At present, most malicious code detection methods are based on the shallow machine learning model. These shallow machine learning methods are simple in the modeling process, and restrict the complex functions and classification problems. In order to improve the accuracy of Andriod malicious code detection and recognition, an algorithm of malicious code detection and recognition of the deep learning has been put forward in this paper, this algorithm based on neural network training and learning model. Through the learning and training of malicious code sample data, the static, dynamic characteristics and malicious application characteristics of malicious code data are analyzed, including privilege feature, API feature and OpCodes characteristic data. The comprehensive performance of the algorithm was tested, the test results indicate that using depth learning detection algorithm, Andriod malicious code identification accuracy is higher, and false detection rate and undetected rate are low, which is a highly efficient and reliable malicious code detection and recognition algorithm.

Copyright

© 2018, 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  - Jing Yang
PY  - 2018/01
DA  - 2018/01
TI  - Android malicious code detection and recognition based on depth learning
BT  - Proceedings of the 2017 4th International Conference on Machinery, Materials and Computer (MACMC 2017)
PB  - Atlantis Press
SP  - 179
EP  - 183
SN  - 2352-5401
UR  - https://doi.org/10.2991/macmc-17.2018.40
DO  - 10.2991/macmc-17.2018.40
ID  - Yang2018/01
ER  -