Malware Detection in Android Systems Using Deep Learning Techniques
Vedavyas R. Niveditha, Santhiya Parivallal, Maria Jones, Amandeep Singh K., P. Rajasekar · Advances in computational intelligence and robotics book series · 2023
Due to the smoothness and various other characteristics, the Android OS is familiar among all kinds of mobile users. Traditionally signature-based techniques are applied to identify malware. But this technique is not able to identify the latest malware. Classification algorithms can support huge datasets needed to protect Android-based platforms. At the same time, a huge dataset needs scalability for detecting and classifying automatically at the malicious identification stage and feature retrieval. In this chapter, enhanced CNN (ECNN) classifier is used for identifying malware in smart devices. The outcome of this suggested classifier is compared with the existing models like XGBoost, random forest, and CNN. The performance of the proposed work is assessed based on their accuracy, precision, and recall values. From the results it is proved that proposed enhanced CNN (ECNN) produces accuracy of 95.8%, precision of 0.96, and recall of 0.92, which is high compared to other algorithms. The tool used for execution is python.