Harmonizing Detection: A Fusion Approach to Combat Adware Attacks on Android Devices

The aim of this paper is, therefore, "the detection of harmful adverts that might be used by attackers to gain unauthorized access in order to mitigate them. This follows that, to minimize this risk, intrusion detection is applied to the combination- based learning technique that is effective in identifying any form of intrusion, called RNN-SVM. Another hybrid solution, RNN-SVM is the combination of Recurrent Neural Networks (RNNs) and Support Vector Machines (SVMs), capable of leveraging the power of these two methods to help outpouring toward improved modern intrusion detection system capabilities at better accuracy and efficiencies. This hybrid approach brings learning skills from Recurrent Neural Networks (RNNs) and the robustness of Support Vector Machines (SVMs) together to enhance the identification of adware attacks over Android-based smartphones and tablets. The testing strategy uses the dataset CCCS_CIC_AndMal2022, which gives a full set of online traffic data that refers to the Adware attack cases in Android devices. Such data hence help the researchers derive effective ways of training their detection systems to test for security breaches. The research, therefore, shows that RNN-SVM technology has the capability of unmasking fraudulent advertising effectively and revealing the efforts of cybercriminals to penetrate Android devices. This further underscores the very reason that advanced machine learning (ML) techniques should be in use to further shore up the security of Android OS devices against adware attacks.