Classifying the malware application in the Android-based smart phones using ensemble-ANFIS algorithm

Nowadays, the Android-based smartphones are fastest gaining in the market to date. Due to its open architecture and ease of application programming interfaces (APIs), it becomes fertile dregs for hackers to deploy malware application. This result in the burglary of personal information those are stored in smartphones, without the user knowledge unauthorised sends unintentional short message, and if the infected smart phones operate remotely it leads ways to some other malware attacks. However, many defence mechanisms were introduced against Android malware; it results in inaccuracy of classification. The contribution of this paper to detect and classify the malwares in the manifest file based on ensemble adaptive neuro-fuzzy inference system (ANFIS) technique. This proposed system is divided into three main steps to detect the malware applications they are: 1) features are extraction using the method called principal component analysis (PCA) method; 2) feature selection, using Pearson correlation coefficient (PCC) method; 3) malware applications are classified, using ensemble of ANFIS technique. The proposed system produced the best detecting malware applications classification and accuracy will be highly efficient than the other classification techniques.