sis 22(1): e7

Research Article

Malware detection for Android application using Aquila optimizer and Hybrid LSTM-SVM classifier

Download116 downloads
  • @ARTICLE{10.4108/eetsis.v9i4.2565,
        author={M. Grace and M. Sughasiny},
        title={Malware detection for Android application using Aquila optimizer and Hybrid LSTM-SVM classifier},
        journal={EAI Endorsed Transactions on Scalable Information Systems},
        volume={10},
        number={1},
        publisher={EAI},
        journal_a={SIS},
        year={2022},
        month={8},
        keywords={Malware detection, Hybrid LSTM-SVM, Aquila optimizer, k-fold Cross validation},
        doi={10.4108/eetsis.v9i4.2565}
    }
    
  • M. Grace
    M. Sughasiny
    Year: 2022
    Malware detection for Android application using Aquila optimizer and Hybrid LSTM-SVM classifier
    SIS
    EAI
    DOI: 10.4108/eetsis.v9i4.2565
M. Grace1,*, M. Sughasiny1
  • 1: Bharathidasan University
*Contact email: grace.phd.2021@gmail.com

Abstract

INTRODUCTION: Android OS is the most recent used smartphone platform in the world that occupies about 80% in share market. In google play store, there are 3.48 million apps available for downloading. Unfortunately, the growth rate of malicious apps in google play store and third party app store has become a big concern, which holds back the development of the Android smartphone ecosystem. OBJECTIVES: In recent survey, a new malicious app has been introduced for every 10 seconds. These malicious apps are built to accomplish a variety of threats, such as Trojans, worms, exploits, and viruses. To overcome this issue, a new efficient and effective approach of malware detection for android application using Aquila optimizer and Hybrid LSTM-SVM classifier is designed. METHODS: In this paper, the optimal features are selected from the CSV file based on the prediction accuracy by cross validation using Aquila optimizer and the mean square error (MSE) obtained by the cross validation is consider as the fitness function for the Aquila to select the optimal features. RESULTS: The extracted optimal features are given to the Hybrid LSTM-SVM classifier for training and testing the features to predict the malware type in the android system. CONCLUSION: This proposed model is implemented on python 3.8 for performance metrics such as accuracy, precision, execution time, error, etc. The acquired accuracy for the proposed model is 97%, which is greater compared to the existing techniques such as LSTM, SVM, RF and NB. Thus, the proposed model instantly predicts the malware from the android application.