IMPLEMENTASI METODE ISOLATION FOREST UNTUK DETEKSI ANOMALI DALAM DATA JARINGAN

Rafli Ardiansyah; Lukman Sunardi; A. Taqwa Martadinata

[1] A. Zulfikar, F. A. Rahmani, N. Azizah, D. J. Perbendaharaan, K. Keuangan, and P. Pinang, “Deteksi Anomali Menggunakan Isolation Forest Belanja Barang Persediaan Konsumsi Pada Satuan Kerja Kepolisian Republik Indonesia,” J. Manaj. Perbendaharaan, vol. 4, no. 1, pp. 1–15, 2023, doi: 10.33105/jmp.v4i1.435.
[2] Baiq Nurul Azmi, Arief Hermawan, and Donny Avianto, “Analisis Pengaruh Komposisi Data Training dan Data Testing pada Penggunaan PCA dan Algoritma Decision Tree untuk Klasifikasi Penderita Penyakit Liver,” JTIM J. Teknol. Inf. dan Multimed., vol. 4, no. 4, pp. 281–290, 2023, doi: 10.35746/jtim.v4i4.298.
[3] G. Martha, “PENGEMBANGAN ALGORITMA MACHINE LEARNING,” vol. 7, pp. 1361–1368, 2024.
[4] S. Situmorang, U. Islam, N. Sumatera, P. Matematika, U. Islam, and N. Sumatera, “Analisis Kinerja Algoritma Machine Learning Dalam Deteksi Anomali Jaringan Sintia Situmorang,” vol. 1, no. 4, 2023.
[5] Milka Wijayanti Sunarto, Dendy Kurniawan, Edy Siswanto, and Haris Ihsanil Huda, “Deteksi Anomali Menggunakan Extended Isolation Forest (Eif),” Tek. J. Ilmu Tek. dan Inform., vol. 1, no. 2, pp. 96–111, 2023, doi: 10.51903/teknik.v1i2.324.
[6] A. Devia and B. Soewito, “Analisis Perbandingan Metode Seleksi Fitur untuk Mendeteksi Anomali pada Dataset CIC-IDS-2018,” J. Teknol. Dan Sist. Inf. Bisnis-JTEKSIS, vol. 5, no. 4, p. 572, 2023, [Online]. Available: http://jurnal.unidha.ac.id/index.php/jteksis
[7] R. M. Imam, P. Sukarno, and M. A. Nugroho, “Deteksi Anomali Jaringan Menggunakan Hybrid Algorithm,” 2019.
[8] D. Firdaus and R. Rianti, “DALAM LALU LINTAS JARINGAN MENGGUNAKAN NAIVE,” vol. 05, no. 02, pp. 140–148, 2023.

Abstract viewed - 22 times
PDF downloaded - 20 times

Affiliations

Rafli Ardiansyah
Universitas Bina Insan

Lukman Sunardi
Universitas Bina Insan

A. Taqwa Martadinata
Universitas Bina Insan

Content Top

IMPLEMENTASI METODE ISOLATION FOREST UNTUK DETEKSI ANOMALI DALAM DATA JARINGAN

Rafli Ardiansyah ,
Lukman Sunardi ,
A. Taqwa Martadinata

Proceedings Economic, Social Science, Computer, Agriculture and Fisheries (ESCAF) 4th 2025

Submitted: Jan 31, 2025

Abstract

Network security is a crucial aspect in the digital age, especially in protecting data and systems from various increasingly complex cyber threats. These threats can not only damage data integrity, but also significantly disrupt organizational operations. This study aims to examine the application of the Isolation Forest algorithm in detecting anomalies in network data. This algorithm was chosen because of its superior ability to efficiently process large-scale data and detect anomalous patterns quickly and accurately. This research covers several important stages, starting from data collection using the Wireshark application. This application generates network data that contains parameters such as packet length, protocol, and source and destination addresses. The data obtained then goes through a preprocessing stage, which includes data cleaning to reduce noise, normalization to equalize the scale between numerical values, and feature selection to select relevant attributes. Once this process is complete, the data is labeled to separate normal and anomalous data, before finally the Isolation Forest algorithm is applied to analyze and detect anomalous patterns. The model that has been developed is evaluated using metrics such as accuracy, precision, recall, and F1- score, which are analyzed through the Confusion Matrix. The results show that the Isolation Forest algorithm is able to detect anomalies with an accuracy of up to 81% at a training and test data sharing ratio of 80:20. This proves the effectiveness of the algorithm in isolating anomalies and maintaining their efficiency despite working with large datasets. This research has contributed significantly to the development of machine learning-based anomaly detection technology. In addition to enriching the literature in this field, this research is also a guide for other practitioners and researchers in improving network security. With the findings obtained, this research is expected to encourage the development of a more effective and efficient network security system in the future.