Permissions-based detection of Malware for android environment using CNN model Available to Purchase

Malware which targets wireless PDAs or mobile phones with the intention of crashing the system as well as leaking or destroying confidential information is known as mobile malware. It is getting more and harder to protect PDA networks and cordless phones from intrusions via viruses as well as other malware as these devices grow more widely used and complex. Users think highly of the growing number of Recently developed malware. Therefore, a new model was proposed for securing the operating system of the Android device by allowing to installed the application or not. The model was based on a dataset of malware for Android applications by building a convolutional neural network CNN and K-NN models for classification stage, whether they are malicious application or not. First apply features select by correlation-based feature selection, then train the applications based on their selected features. In order to test the accuracy of two models, Confusion Matrix is used. The last stage, after detecting the malware is let the user to allow or not to activate the download of the application to the Android device. The entire model is proposed to provide a secure environment for Android from malware. It has been proven that the permissions-based approach can classify malware or regular software with high accuracy using the CNN model, reaching 0.9880, while the accuracy value was 0.9711 using the KNN.