Android Malware Detection Model using the Adaptive Swarm Optimization-based Neural Network
Pallavi Madhukarrao Shimpi, Nitin Namdeo Pise · 2023
In industries and transportation systems, Android-based applications and devices were extensively utilized and deployed. Some of the most efficient and successful methods for ensuring the security of Android systems, particularly for commercial platforms and smart cities, is malware detection. Furthermore, there has been a significant increase in research employing deep learning-based methods to identify malicious activity. Hence, in this research, the adaptive swarm optimization (ASO) technique is proposed for effective android malware detection by their movement and the intelligent behavior of the swarm. These significant characters are incorporated into the utilized neural network (NN), which is preferred to detect the event as malware or not. The ASO model assists to regulate the hyperparameters of the utilized classifier, in which the performance is enhanced for malware detection. The ASO-based NN attains an accuracy of 98.417 %, a sensitivity of 98.943 %, and a specificity of 97.824 % in android-based malware attack detection.