Application of machine learning in malware detection for Android Available to Purchase

The fight to reduce Android malware has been seen as an important endeavour as the use of smart devices powered by the Android operating system grows in popularity complex evasive methods that require more effort modern methods of detection. Consequently, in this article, two Support Vector Machine (SVM) techniques for Machine Learning (ML) K-Nearest neighbours (KNN) and Support Vector Machine (SVM) are used and assessed to carry out the feature set’s categorization into applications (apps), whether good or bad, via guided education procedure. Static app analysis is part of this study. which examines if keywords are present and used often in the static feature sets are derived from the Android applications’ manifest file. a collection of 400 apps will help detect malware more accurately. The experimental findings show that employing SVM and KNN, the average accuracy rates for a dataset of genuine malware and benign applications are 88.66 percent and 80.33 percent, respectively, with average true positive rates of over 70.0 percent and 79.66 percent.