Please use this identifier to cite or link to this item: https://repository.uksw.edu//handle/123456789/37292
Title: Klasifikasi Serangan pada Jaringan Komputer berbasis Ensemble Learning menggunakan Gradient Boosting Classifier dan AdaBoost
Other Titles: -
Authors: Lamalengga, Ari Crismast
Keywords: intrusion detection system;machine learning;ensemble learning;gradient boosting
Issue Date: 14-May-2025
Abstract: Penelitian ini menggunakan Gradient Boosting dan AdaBoost untuk klasifikasi serangan pada Intrusion Detection System (IDS). Training dan testing dilakukan pada dataset UNSW-NB15. Klasifikasi yang dilakukan adalah binary classification di mana classifier tersebut bekerja untuk menentukan apakah sample data yang diproses dari testing dataset merupakan serangan (attack) atau bukan serangan (normal). Hasil pengujian menunjukkan Gradient Boosting dengan jumlah estimator 150 memperoleh nilai detection rate sebesar 99.40% dan akurasi sebesar 82.47%. AdaBoost dengan jumlah estimator 100 memperoleh nilai detection rate 99.88% dan nilai akurasi 80.82%.
This study uses Gradient Boosting and AdaBoost for attack classification on the Intrusion Detection System (IDS). Training and testing were conducted on the UNSW-NB15 dataset. The classification performed was binary classification, where the classifier worked to determine whether the data sample processed from the testing dataset was an attack or not an attack (normal). The test results showed that Gradient Boosting with 150 estimators obtained a detection rate of 99.96% and an accuracy of 81.02%. AdaBoost with 100 estimators obtained a detection rate of 99.88% and an accuracy value of 80.82%.
URI: https://repository.uksw.edu//handle/123456789/37292
Appears in Collections:T1 - Informatics Engineering

Files in This Item:
File Description SizeFormat 
T1_672019180_Judul.pdf696.68 kBAdobe PDFView/Open
T1_672019180_Isi.pdf
  Until 9999-01-01
470.28 kBAdobe PDFView/Open
T1_672019180_Daftar Pustaka.pdf445.35 kBAdobe PDFView/Open
T1_672019180_Lisensi dan Embargo.pdf
  Until 9999-01-01
362.15 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.