Malware Detection Technique for Android Devices Using Machine Learning Algorithms

The increasing popularity of Android devices has raised concerns about the growing risk of malware. Malicious apps can compromise data security, disrupt device functionality, and invade user privacy. Traditional methods of detecting malware based on signatures need help to keep up with the evolving software landscape. This study introduces an approach to identifying Android malware using machine learning (ML) techniques. By extracting features of Android apps such as permissions requested by Android applications the method trains ML models to distinguish between benign and malicious applications. The effectiveness of these models is ensured using a comprehensive dataset. The paper also explores the performance of ML algorithms like Logistic Regression (LR), Support Vector Machines (SVM), Decision Trees (DT), Random Forest (RF), k-Nearest Neighbors (k-NN), Neural Networks (NN), Gradient Boosting Machines (GBM), Naive Bayes (NB), AdaBoost (AdaB), XGBoost (XGB) in detecting malware. Practical aspects like efficiency and scalability are discussed, emphasizing how these methods can seamlessly integrate into Android security solutions. This research represents a step in detecting Android malware offering improved protection against emerging security risks for Android users. The results show that RF obtained the highest accuracy of 97.20<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> using the NATICUSdroid (Android Permission) Dataset for malware detection for Android devices.