Android Malware Family Classification using Ensembling of Fpt and Fcm with Decision Tree

Main Article Content

Raju Kumar Ranjan, Manoj Sethi

Abstract

Android malware classification and assigning the appropriate android malware family is challenging. Traditional static analysis methods can easily be misguided by malware, and dynamic analysis consumes more space and time. This research proposed a fuzzy-based android malware family classification using multiple aspects of the DEX file. The considered aspects are Permissions of Android application, Image obtained from DEX file sectional features, Dalvik Opcode, and Bytecode of corresponding DEX file. The feature vectors acquired from these multiple aspects are fuzzified using a triangular fuzzifier. The obtained fuzzy sets are classified using an FPT classifier and clustered using Fuzzy C-means. FPT and FCM are combined according to the views, and a Decision Tree model is obtained for classifying the Android malware family. The final model produces an accuracy of up to 95.75%.

Article Details

Section
Articles