Dynamic Ensemble Learning Framework Enhanced with XAI To Detect Android Malware

The unparalleled threat posed by Android malware is the primary cause of numerous security issues on the internet. Although commendable efforts have been made to detect and classify Android malware using machine learning techniques, more dynamic learning must be explored for this purpose. The cyber community must address the detection of Android malware in smartphones to eliminate these menacing malware samples. This study presents a new approach that utilizes dynamic ensemble learning with enhanced XAI to classify and characterize Android malware samples. The CIC-AndMal2020 dataset used in this study includes 14 prominent malware categories and 191 distinguished malware families. This research demonstrates the successful application of dynamic ensemble learning, achieving an accuracy of 98.36% and the efficiency of the results is confirmed with explainable AI.