Machine Learning based Multilayered Feature Analysis for Android Malware Detection

Malicious applications have grown exponentially as a result of the widespread usage of smartphones running Android, endangering user privacy and security. Conventional malware detection methods, which are frequently signature-based, find it difficult to keep up with the quickly changing malware landscape. A strong machine learning-based detection method that uses multilayered feature analysis to improve Android malware detection is presented in this study. The suggested methodology combines network-level, dynamic, and static feature analysis to completely identify a variety of harmful application behavioral patterns. The performance of five machine learning algorithms Random Forest (RF), Gradient Boosting, Support Vector Machine (SVM), K-Nearest Neighbors (KNN) and Naive Bayes on the AndroZoo dataset is evaluated. Tree-based models like RF and Gradient Boosting perform better than other classifiers, as shown by numerous studies, with accuracy levels above 94% and high precision, recall, and F1-score values. The multilayered feature analysis approach highlights important traits that contribute to malware categorization, improving both the interpretability of model decisions and detection accuracy.