A Study on Clustering Classification Technique based on Machine Learning to Detect Android Malware Variants

Authors

  • Woong Go Korea Internet & Security Agency
  • Jun-hyung Park Korea Internet & Security Agency

Keywords:

Clustering; Machine Learning; Malware Variants; Detection, Classification

Abstract

Mobile malware found these days are distributed for financial gain. Most of those malware are created and
used as a malware variant that re-uses existing malicious behavior, because the financial objective can be achieved
efficiently at low cost, compared with creating new malware. Another reason is that mobile malware with a short life
cycle can be created massively to spread infection. However, anti-malware solutions available these days detect malware
using the known signature of malware. Therefore, those solutions have a limit in detecting a malware variant that
modifies existing malware partially. If many malware variants can be detected quickly, infection spread can be blocked
in early stages and damages can be reduced. This paper proposes a clustering classification technique based on the
unsupervised machine learning algorithm, which is designed to detect malware variants quickly that seek financial gain.

Published

2018-01-25

How to Cite

Woong Go, & Jun-hyung Park. (2018). A Study on Clustering Classification Technique based on Machine Learning to Detect Android Malware Variants. International Journal of Advance Engineering and Research Development (IJAERD), 5(1), 903–908. Retrieved from https://ijaerd.org/index.php/IJAERD/article/view/2204