Abstract
There has been an exponential rise in the development of Android applications, leading to an escalated vulnerability of mobile gadgets against malicious code that compromises privacy and system integrity. However, traditional malware detection methods have been dependent upon high-dimensional feature learning, making them less efficient and deployable. This article introduces the concept of an effective framework for Android malware detection that uses Genetic Algorithm-based static feature optimization techniques for improving classifying efficiency and simplifying the model. It makes use of static attributes such as the permissions and APIs used by an Android app extracted by parsing Android app packages, thereby applying an evolutionary approach for static feature optimization. The optimal set is used for training classifiers such as Support Vector Machines and Artificial Neural Networks. Simulation results show that the proposed framework performs well with improved efficiency and lower-dimensional learning capabilities. It was observed that GA-based static learning for malware detection in Android offers an effective technique that is fast and scalable.