Android malware classification using whale optimization algorithm

Abstract

The accuracy of any classification algorithm essentially depends on the cohesiveness and structure of the training dataset and its features. The detection of malicious applications running on android devices has become a task that cannot be overemphasized. This is due to the wide acceptability and usefulness of these devices. This usefulness however has also made the Android applications to become soft targets for malware hackers. In order to ameliorate this problem, different malware detection techniques have been proposed in the literature. However, the accuracy and false alarm rate still require improvements in order to have a versatile detector. This research therefore presents the use of Whale Optimization Technique for feature selection of permission-based feature of Android applications for better classification accuracy. The results show that the accuracy is improved using this algorithm compare to some known existing detector models with or without feature selector.