Detecting Malicious Threats On Mobile Applications Using Machine Learning Approaches

In Today scenario , Android is one of the most frequently used mobile operatings systems, thus it is a priority for advanced threat actors and hackers. Malicious code is often found in Android applications to which security solutions have a problem of identifying the threats. Today’s Android malware is sophisticated enough that it can avoid most detection .Therefore, machine learning is more efficient in stopping a large number of diverse contemporary threats to Android devices. These machine learning methods essentially first rely on the detection of sequential patterns of malicious behaviors, and secondly, distinguish between existing and new types of threats that have previously exhibited typical and atypical behavior. Our proposed approach builds on Ensemble learning where the methods include Support Vector Machines, Decision Trees, Logistic regression and XGBoost to bolster the detection abilities of the model. This methodology serves to enhance the specific model performance on identifying known malware .