Design of Machine Learning-Based Malware Detection Techniques in Smartphone Environment

Abstract

Hackers are spreading malware by using Android mobile devices and apps. Malware detection in Android apps is a topic of study. How can we use ML to identify harmful software in Android-based gadgets and programs? If Android apps could detect malware in real-time, it may better protect users from it. To put it simply, it will assist Android users to avoid downloading harmful apps. To aid in supervised learning, the suggested technique gathers features from APK files. Multinomial Naive Bayes, Random Forest, and Support Vector Machines are only a few of the prediction models available (SVM). To counteract harmful malware, Android devices and apps may rely on a foundation created by ML methods. There is more backing for the solution proposed. More and more malware is being discovered, and hence more training data is being collected. When more data is used for training, accuracy improves. Small tweaks might make it possible to live-track Android apps on rival devices.