Year: 2025
Type: article
Abstract: Android devices are becoming increasingly popular, and there are more threats to Android users. This paper discusses Android malware detection using a deep learning classification approach. In this study, Android software was analysed using malware analysis tools, the selected features were extracted, and the results were compiled into a CSV file. Then, its use in CNN and RNN models for malware detection was analysed by measuring accuracy using the standard accuracy formula. According to the development process, CNN performs better at detecting Android malware, achieving 96 per cent accuracy, while RNN achieves 75 per cent accuracy.
Institutions: Universiti Teknologi MARA, Universiti Teknologi MARA System
Language: English
Subfield: Software
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7128807796
