Metaheuristics with Deep Learning Model for Cybersecurity and Android Malware Detection and Classification

Since the development of information systems during last decade, cybersecurity has become a critical concern for many groups, organizations, and institutions. Malware applications are among commonly used tools tactics perpetrating cyberattack on Android devices, it is becoming challenging task to develop novel ways identifying them. There various malware detection models available strengthen operating system against such attacks. These detectors categorize target based patterns that exist in features present applications. As analytics data continue grow, they negatively affect defense mechanisms. large numbers unwanted create performance bottleneck mechanism, feature selection techniques found be beneficial. This work presents Rock Hyrax Swarm Optimization with deep learning-based (RHSODL-AMD) model. The technique presented includes finding Application Programming Interfaces (API) calls most significant permissions, which results effective discrimination between good ware Therefore, an RHSO subset (RHSO-FS) derived improve classification results. In addition, Adamax optimizer attention recurrent autoencoder (ARAE) model employed detection. experimental validation RHSODL-AMD Andro-AutoPsy dataset exhibits its promising performance, maximum accuracy 99.05%.