Android Malware Detection

With the rapid expansion of Android-based mobile applications, ensuring user security and privacy has become a growing concern. Android's open-source nature and widespread adoption have made it a prime target for malware developers. Traditional malware detection approaches, such as signature-based and heuristic techniques, are increasingly insufficient against sophisticated and evolving threats. This project aims to develop an intelligent Android malware detection system using machine learning techniques to identify malicious applications based on behavioral and static features extracted from APK files. The proposed system utilizes a comprehensive dataset consisting of both benign and malicious Android applications. Static analysis is performed by extracting features such as permissions, API calls, intents, and hardware components, without executing the code. These features are then preprocessed and used to train multiple machine learning models including Random Forest, Support Vector Machine (SVM), and Gradient Boosting Classifier. Among these, the Random Forest classifier demonstrated superior accuracy, precision, and recall in detecting malware, making it the most suitable model for deployment.