Analisis Deteksi Malware Android Menggunakan Metode Support Vector Machine & Random Forest

Authors

DOI:

https://doi.org/10.34818/eoe.v8i6.16864

Abstract

Abstrak Menurut data laporan kementerian komunikasi dan informatika, tahun 2018 pengguna aktif smartphone di Indonesia lebih dari 100 juta orang serta pada tahun yang sama data dari statcounter pengguna android di Indonesia sebanyak 90,85%. Tingginya penggunaan android membuat sistem operasi android menjadi target serangan malware. Malware merupakan sebuah sistem yang diprogram agar dapat menyusup ke sebuah sistem operasi, sebuah sistem operasi yang telah diserang malware dapat mengalami kerusakan dan bahkan dengan niat yang lebih jahat malware dapat digunakan untuk mencuri data-data penting. Rentannya serangan malware dan dapat merugikan para pengguna android sehingga diperlukan analisis lebih lanjut, oleh karena permasalah yang ada mendorong penelitian ini dilakukan dengan tujuan untuk deteksi dini. Pada kasus ini digunakan pendekatan machine learning untuk melakukan klasifikasi data serangan malware android. Machine learning yang digunakan dalam penelitian ini adalah Support Vector Machine (SVM) dan Random Forest. Kedua metode machine learning itu dipilih karena pada penelitianpenelitian sebelumnya terbukti kedua metode itu sangatlah efektif melakukan klasifikasi dengan menghasilkan akurasi yang tinggi. Pada makalah ini dilakukan perbandingan antara metode Support Vector Machine (SVM) dengan metode Random Forest dalam melakukan klasifikasi data, serta membandingkan hasil akurasi dengan penelitian sebelumnya. Proses klasifikasi menggunakan metode Support Vector Machine (SVM) menghasilkan nilai precision 97%, nilai recall 97%, dan nilai f1-score 97%, dan akurasi 96,23%, pada metode Random Forest menghasilkan nilai precision 99%, nilai recall 99%,nilai f1-score 99%, dan akurasi 98,99%. Menurut hasil percobaan, metode Random Forest lebih unggul dari metode Support Vector Machine (SVM) dan pendekatan yang diusulkan di penelitian ini memiliki hasil performansi matrik di atas 95% yang lebih baik daripada penelitian sebelumnya. Kata kunci : Android, Malware, Machine learning, Support Vector Machine (SVM), Random Forest. Abstract According to data from the ministry of communication and informatics, in 2018 active smartphone users in Indonesia more than 100 million people and in the same year data from the android user statcounter in Indonesia as much as 90.85%. The high use of android makes the android operating system a target for malware attacks. Malware is a system that is programmed to infiltrate an operating system, an operating system that has been attacked by malware can be damaged and even with more malicious intentions malware can be used to steal important data. The vulnerability of malware attacks and can harm android users so further analysis is needed, because the existing problems encourage this research to be done with the aim of early detection. In this case, a machine learning approach is used to classify android malware attack data. The machine learning used in this study is support vector machine (SVM) and random forest. Both machine learning methods were chosen because in previous studies it was proven that both methods are very effective at classifying with high accuracy. In this paper, a comparison between the Support Vector Machine (SVM) method and the Random Forest method in classifying data, as well as comparing the results of accuracy with previous research. The classification process using the Support Vector Machine (SVM) method produces a precision value of 97%, a recall value of 97%, and an f1-score value of 97%, and an accuracy of 96.23%, in the Random Forest method it produces a precision value of 99%, a recall value of 99%, an f1-score value of 99%, and an accuracy of 98.99%. According to the results of the experiment, the Random Forest method is superior to the Support Vector Machine (SVM) method and the approach proposed in this study has a higher matric performance result above 95% than previous studies. Keywords : Android, Malware, Machine learning, Support Vector Machine (SVM), Random Forest.

References

[1] Egham, 2017,”Gerter Says Worldwide Sales of Smartphone Grew 9 Percent in First Quarter of 2017”,

https://www.gartner.com/newsroom/id/3725117 akses: Juli 2021

[2] Kementerian Komunikasi dan Informatika, 2018, “Indonesia Raksasa Teknologi Digital Asia”,

https://kominfo.go.id/content/detail/6095/indonesia-raksasa-teknologi-digital-asia/0/sorotan_media

akse: Juli 2021

[3] Startcounter Global Stats, 2018, “Mobile Operating System Market Share Indonesia”,

https://gs.statcounter.com/os-market-share/mobile/indonesia/2018 , akses: Juli 2021

[4] Imantoko Kurniadi, 2021, “Serangan Mobile Malware terdeteksi Menurun di Indonesia”

https://selular.id/2021/04/serangan-mobile-malware-terdeteksi-menurun-di-indonesia/ , akses: Juli 2021

[5] C. C. U. Lopez, A. N. Cadavid, " Framework for malware analysis in Android",.i2t Research Group , Icesi

University, Cali, Colombia, 2016.

[6] D. O. Sahin, O. E. Kural, S. Akleylek, E. Kilic, “New Result on Permission Based Static Analysis For

Android Malware”, Department of Computer Engineering , Ondokuz Mayis University, 2018.

[7] T. Ban, T. Takahashi, S. Guo, D. Inoue, K Nakao, “Integration of Multi-modal Features For Android

Malware Detection Using Linear SVM”, Information and Communication Tachnology, Koganei, Tokyo,

2016.

[8] H. J. Zhu, T.H. Jiang, Bo Ma, Z.H. You, W.L. Shi, Li Cheng, “HEMD: a highly efficient random forest

based malware detection framework for Android”, The Xianjiang Technical Institute of Pyisics and

Chemistry, Chinese Academy of Sciences, China, 2016.

[9] Boser B., Guyon I., Vapnik V., “Proceedings of the fifth annual workshop on Computational learning

theory”, COLT, 1992

[10] Ho Tin Kam, “Random Decision Forests”, AT&T Bell Laboratories 600 Moutain Avenue, Murray Hill,

USA , 1995

[11] https://www.kaggle.com/xwolf12/datasetandroidpermissions akses: Juni 2020

[12] Leo Breiman, “Random Forests”, Statis Department, University of California, Berkeley, 2001.

[13] Junmei S., Kai Y., X Liu , Chunlei Y., Yaoyin F., “Malware Detection on Android Smarthone using

Keywords Vector and SVM ”, Hangzhou Institute of Sorvice Engineering, Hangzhou Normal University,

China, 2017.

[14] Yuki T., Kazuya S., Satoshi F., “Detecting Ransomware using Support Vector Machines”, Department of

Electrical Engineering and Computer Science, Tokyo Metropolita University, Japan, 2018.

[15] Nikola M., Ali D., Kim-Kwang R. C., “Machine learning aided Android malware classification”, School of

Computer Science, University of Manchester, UK, 2017.

[16] Alif P. B.A, Yudha P., M. Faris R., “Deteksi Malware Menggunakan Metode Random Forest Berdasarkan

Analisa Forensik”, Prodi S1 Teknik Komputer, Universitas Telkom, 2021.

[17] Ary Adhigana S., Parma S., Erwid M. J., “Analisis Metode Ensemble untuk Mendeteksi Malware pada

Mobile Devices”. Fakultas Informatika, Universitas Telkom, 2019.

[18] Mahendra D. P., M. Irvan T., Adnan I. H., Adiwijaya, “Perbandingan Support Vector Machine dan

Modified Balanced Random Forest dalam Deteksi Penyakit Diabetes”, Rekayasa Sistem dan Teknologi

Infomasi, Universitas Telkom, 2019.

[19] https://developer.android.com/reference/android/Manifest.permission

[20] http://androidpermissions.com/

Downloads

Published

2021-12-01

Issue

Pages

12500-12518

Section

Program Studi S1 Rekayasa Perangkat Lunak