Treffer: A novel hybrid machine learning approach for analysis and detection of malicious mobile applications.

This work creates a new framework based on permission-based analysis for Android malware detection. This collection uses static analysis to extract program permissions, taking advantage of the crucial role that permissions play in Android security. It audits the safety of applications using machine learning approaches like multiple linear regression. In general, two classifiers for the detection of permission-based Android malware are suggested; they are both based on linear regression models. Interestingly, these models demonstrate remarkable performance, eliminating the need for excessively complex categorisation techniques. These linear regression-based classifiers provide unbelievable performance in Android malware identification when compared to Decision Tree Classifier, Naïve Bayes, KNN, and Support Vector Machines. [ABSTRACT FROM AUTHOR]