HamDroid: permission-based harmful android anti-malware detection using neural networks

Saeed Seraj, Siavash Khodambashi, Michalis Pavlidis, Nikolaos Polatidis

Research output: Contribution to journalArticlepeer-review

Abstract

Android platforms are a popular target for attackers, while many users around the world are victims of Android malwares threatening their private information. Numerous Android anti-malware applications are fake and do not work as advertised because they have been developed either by amateur programmers or by software companies that are not focused on the security aspects of the business. Such applications usually ask for and generally receive non-necessary permissions which at the end collect sensitive information. The rapidly developing fake anti-malware is a serious problem, and there is a need for detection of harmful Android anti-malware. This article delivers a dataset of Android anti-malware, including malicious or benign, and a customized multilayer perceptron neural network that is being used to detect anti-malware based on the permissions of the applications. The results show that the proposed method can detect with very high accuracy fake anti-malware, while it outperforms other standard classifiers in terms of accuracy, precision, and recall.

Original languageEnglish
Pages (from-to)15165-15174
Number of pages10
JournalNeural Computing and Applications
Volume34
Issue number18
DOIs
Publication statusPublished - 13 Jan 2022

Bibliographical note

Publisher Copyright:
© 2021, The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature.

Keywords

  • Android
  • Fake anti-malware
  • Malware detection
  • Neural networks

Fingerprint

Dive into the research topics of 'HamDroid: permission-based harmful android anti-malware detection using neural networks'. Together they form a unique fingerprint.

Cite this