DETECTION OF MOBILE MALWARES USING IMPROVED DEEP CONVOLUTIONAL NEURAL NETWORK
D. Hemalatha · Journal of Critical Reviews · 2020
Android, the leading mobile operating system (OS), has since its launch in 2008 become a robust smart device platform. Android has millions of active users, which allow hackers and cyber criminals to make malware added to the platform. In the past several years, a variety of security measures were developed and implemented on Android malware analyzers and detectors, including a unique ID (UID) for each device, system authorization and Google Play distribution platform. The modern malware and detection methods for Android have considerably improved, leaving most traditional malware detection methods obsolete. This article provides a method to detect malicious Android malware by using Deep Convolutionary Neural Network (DCNN). This proposed Deep Convolutionary Neural Network (DCNN-RLU) is developing mixed kernel learning algorithms for Android malware detection. The analysis is based on a data set of the MalGenome. Experimental findings indicate that the Deep Neural network system output reached up to 98.769%, suggesting the efficiency of the DCNN-RLU model proposed. In contrast with the current methods, we address the performance gains from the proposed Deep Learning approach.