PENGEMBANGAN MODEL DETEKSI MALWARE PERMISSION-BASED PADA APLIKASI ANDROID MENGGUNAKAN SUPPORT VECTOR MACHINE DENGAN OPTIMASI FUNGSI KERNEL

Bagus Syamsu Rahmatullah, - (2025) PENGEMBANGAN MODEL DETEKSI MALWARE PERMISSION-BASED PADA APLIKASI ANDROID MENGGUNAKAN SUPPORT VECTOR MACHINE DENGAN OPTIMASI FUNGSI KERNEL. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Perkembangan teknologi digital dan tingginya ketergantungan pada perangkat mobile, khususnya Android yang menguasai 79% pasar di Asia, telah memicu peningkatan signifikan dalam ancaman serangan siber, terutama serangan malware. Serangan siber berbasis malware telah menyumbang serangan sebesar 60% dari total 1.637.973.022 serangan yang tercatat pada tahun 2021 di Indonesia, penting untuk memahami mekanisme serangan siber serta metode pencegahannya. Penelitian ini menguraikan pengembangan model deteksi malware permission-based pada aplikasi Android dengan menggunakan algoritma Support Vector Machine (SVM), karena algoritma ini efektif untuk mengklasifikasikan data yang tidak linier dan berdimensi tinggi. Selain itu, penelitian ini bertujuan untuk mengevaluasi efektifitas SVM dalam mendeteksi malware dengan metode permission-based dan mengevaluasi pengaruh fungsi kernel (linear, polynomial, rbf dan sigmoid) terhadap akurasi model deteksi malware. Hasil penelitian menunjukkan bahwa kernel polinomial dan Radial Basis Function (RBF) secara signifikan meningkatkan akurasi deteksi, mencapai hingga 87% sementara kernel sigmoid menunjukkan kinerja yang paling rendah sebesar 56%. Pencapaian ini turut dipengaruhi oleh pemilihan fitur perizinan aplikasi yang relevan, karena fitur tersebut merefleksikan pola akses sistem yang menjadi indikator kuat dalam membedakan aplikasi jinak dan malware. Penelitian ini memberikan kontribusi pada bidang keamanan siber dengan memberikan wawasan tentang penerapan machine learning untuk deteksi malware dengan pendekatan permission-based yang efisien. Implementasi fungsi kernel yang tepat dalam model deteksi malware berbasis SVM secara signifikan meningkatkan akurasi dan efisiensi sistem keamanan siber dalam mendeteksi malware.
---------
The advancement of digital technology and the increasing reliance on mobile devices, particularly Android, which commands 79% of the market in Asia, has led to a significant rise in cyber attack threats, especially those involving malware. Malware-based cyber attacks accounted for 60% of the total 1,637,973,022 recorded attacks in Indonesia in 2021, highlighting the necessity to comprehend the mechanisms of cyber attacks and their prevention methods. This study outlines the development of a permission-based malware detection model for Android applications utilizing the Support Vector Machine (SVM) algorithm, known for its effectiveness in classifying non-linear and high-dimensional data. Furthermore, the research aims to assess the effectiveness of SVM in detecting malware through a permission-based approach and to evaluate the impact of various kernel functions (linear, polynomial, RBF, and sigmoid) on the accuracy of the malware detection model. The findings indicate that both polynomial and RBF kernels significantly enhance detection accuracy, achieving up to 87%, while the sigmoid kernel exhibited the lowest performance at 56%. This achievement is also influenced by the selection of relevant application permission features, as these features reflect system access patterns that serve as strong indicators for distinguishing benign applications from malware. This research contributes to the field of cybersecurity by providing insights into the application of machine learning for efficient malware detection using a permission-based approach. The appropriate implementation of kernel functions within the SVM-based malware detection model significantly improves the accuracy and efficiency of cybersecurity systems in identifying malware.

[thumbnail of S_RPL_2003164_Title.pdf] Text
S_RPL_2003164_Title.pdf

Download (282kB)
[thumbnail of S_RPL_2003164_Chapter1.pdf] Text
S_RPL_2003164_Chapter1.pdf

Download (49kB)
[thumbnail of S_RPL_2003164_Chapter2.pdf] Text
S_RPL_2003164_Chapter2.pdf
Restricted to Staf Perpustakaan

Download (383kB)
[thumbnail of S_RPL_2003164_Chapter3.pdf] Text
S_RPL_2003164_Chapter3.pdf

Download (155kB)
[thumbnail of S_RPL_2003164_Chapter4.pdf] Text
S_RPL_2003164_Chapter4.pdf
Restricted to Staf Perpustakaan

Download (402kB)
[thumbnail of S_RPL_2003164_Chapter5.pdf] Text
S_RPL_2003164_Chapter5.pdf

Download (33kB)
[thumbnail of S_RPL_2003164_Appendix.pdf] Text
S_RPL_2003164_Appendix.pdf
Restricted to Staf Perpustakaan

Download (1MB)
Official URL: https://repository.upi.edu/
Item Type: Thesis (S1)
Additional Information: SINTAID: 6682222 SINTAID: 6681751
Uncontrolled Keywords: Android, Deteksi Malware, Machine Learning, Support Vector Machine, Fungsi Kernel, Malware Detection, Kernel Function
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76 Computer software
T Technology > T Technology (General)
Divisions: UPI Kampus cibiru > S1 Rekayasa Perangkaat Lunak
Depositing User: Bagus Syamsu Rahmatullah
Date Deposited: 09 May 2025 03:58
Last Modified: 09 May 2025 03:58
URI: http://repository.upi.edu/id/eprint/133062

Actions (login required)

View Item View Item