INTRUSION DETECTION MENGGUNAKAN BACKPROPAGATION NEURAL NETWORK UNTUK MENCEGAH PELANGGARAN KEAMANAN PADA SISTEM KOMPUTER

Anillah, Khafi (2020) INTRUSION DETECTION MENGGUNAKAN BACKPROPAGATION NEURAL NETWORK UNTUK MENCEGAH PELANGGARAN KEAMANAN PADA SISTEM KOMPUTER. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
Laporan Tugas Akhir Khafi (1) (1) (2) (1).pdf - Accepted Version
Restricted to Repository staff only

Download (1MB) | Request a copy

Abstract

Jaringan komputer saat ini memiliki peran penting dalam berbagai bidang, dari pemerintahan hingga perusahaan pribadi. Seiring fakta tersebut, muncul banyak pelanggar keamanan jaringan komputer yang bertujuan untuk mendapatkan keuntungan pribadi. Oleh karena itu, tiap komputer disarankan memiliki Intrusion Detection System (IDS). Namun terdapat masalah fundamental dalam IDS, yakni kuantitas peringatan yang terlalu banyak sehingga merepotkan pengguna, dan performa yang terus menurun karena kesalahan pelaporan dan banyaknya alarm palsu pada kondisi normal. Untuk itu diperlukan sistem yang dapat memeriksa pelanggaran secara efektif dan efisien. Artificial Neural Networks (ANN) adalah sebuah metode dari Machines learning yang berkerja seperti sistem saraf pada makhluk hidup, dimana sistem ini akan bereaksi pada ransangan. Tujuan dari diciptakannya ANN adalah agar sistem komputer bekerja layaknya otak manusia, namun lebih cepat dan akurat. Sehingga, metode ini dirasa tepat untuk memecahkan masalah ini. Dengan metode ANN, diperoleh IDS dengan akurasi diatas 80%, nilai recall dan precision diatas 0,5, dan sistem yang mengalami penurunan pada iterasi ke-150. ========================================================= Today's computer networks have an important role in various fields, from government to private companies. Along with this fact, there will be many network security violators that aim to get personal profit from computer networks. Therefore, it is recommended that each computer have an Intrusion Detection System (IDS). However, there are fundamental problems in IDS, namely the quantity of warnings that are too much so that it is troublesome for users, and performance continues to decline due to reporting errors and the number of false alarms under normal conditions. This requires a system that can examine violations effectively and efficiently. Artificial Neural Networks (ANN) is a method of learning machines that work like the nervous system in living things, where the system will react to stimuli. The purpose of the creation of ANN is that computer systems work like the human brain, but are faster and more accurate. Thus, this method is considered appropriate to solve this problem. With the ANN method, IDS is obtained with accuracy above 80%, recall and precision values above 0.5, and the system has decreased at the 150th iteration.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Intrusion Detection System, Artificial Neural Network, Jaringan Komputer
Subjects: Q Science > QA Mathematics
Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User: Khafi Anillah
Date Deposited: 29 Aug 2020 04:27
Last Modified: 29 Aug 2020 04:27
URI: https://repository.its.ac.id/id/eprint/81544

Actions (login required)

View Item View Item