A support vector machine water wave optimization algorithm based prediction model for metamorphic malware detection

Abstract

In this paper, we proposed a novel method based on coupling of SVM (Support Vector Machine) and WWO (Water Wave Optimization) for detection of metamorphic malware. The working of SVM model depends upon the proper selection of SVM parameters. Malware signatures have been taken from G2, MWOR, MPCGEN and NGVCK (Next Generation Virus Creation Kit).Benign signatures have been taken from Gygwin, GCC, TASM, MingW and Clang.ClustalW and T-Coffee are used for signature alignment during primary pairwise alignment and secondary multiple alignment in order to avoid the problem of variable length of code. In this study WWO has been employed for determining the parameters of SVM. The performance of SVM-WWO method has been compared with LAD Tree, Naive Bayes, SVM and ANN(Artificial Neural Network). Furthermore, The results obtained show that the newly proposed approach provides significant accuracy. Satisfactory experimental results show the efficiency of proposed method for metamorphic malware detection. Further, it has been recommended that this method can be used to facilitate commercial antivirus engines.