Comparative Analysis of Genetic Algorithm and Simulated Annealing for Feature Selection in Android Malware Detection Across Multiple Feature Sets

As mobile applications are so widely used, Android malware continues to pose a serious concern. In order to improve malware detection efficiency across multiple static feature sets—hardware components, permissions, intents, and their combinations (hardware-intents, intents-permissions, permissions-hardware, and all three—this paper compares two feature selection techniques: Genetic Algorithm (GA) and Simulated Annealing (SA). For every feature set, the effectiveness of various metaheuristic approaches is contrasted with more conventional approaches such as Principal Component Analysis (PCA), Recursive Feature Elimination (RFE), and a baseline (no feature selection). Accuracy, F1-score, and the quantity of characteristics used are used to assess the results. According to the given feature information, the study shows how well GA and SA find pertinent features for various data modalities, indicating their potential to maximise malware classification performance.