Android Malware Detection Using Feature Selection With Hybrid Genetic Algorithm and Simulated Annealing

The popularity of the android operating system and the easy development of applications on the android platform have made it easy for anyone to produce malware by using prepared tools. this has led to the spread of malware among many useful applications that can cause problems for android users. in this study, we have provided a way to detect android malware by using permissions that have been obtained in the form of static analysis. in the proposed method, we select the relevant features from the set of permission by combining genetic algorithm and simulated annealing, and three algorithms gasa-svm, gasa-dt, and gasa-knn are developed based on this approach. the proposed method is evaluated on a portion of the drebin dataset, which included 410 samples with 82 malware and 328 benign application. the proposed method improves android malware detection accuracy, and the gasa-svm with the best value of 0.9707 has the best result.