Classification Iot Botnet Attacks Using Explainable Artificial Intelligence (Xai) With Decision Tree And Random Forest

Purba, Rahel Cecilia (2024) Classification Iot Botnet Attacks Using Explainable Artificial Intelligence (Xai) With Decision Tree And Random Forest. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025201155-Undergraduate_Theses.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (1MB) | Request a copy

Abstract

Internet of Things atau yang biasa disebut IoT merupakan salah satu bukti kemajuan teknologi. IoT memberikan banyak manfaat bagi penggunanya dan menawarkan kemudahan dalam kehidupan sehari-hari. Namun, IoT tidak hanya memberikan manfaat, tetapi juga tantangan bagi penggunanya. Tantangan tersebut adalah adanya aktivitas mencurigakan di jaringan lalu lintas IoT. Aktivitas mencurigakan tersebut sering disebut sebagai botnet, IoT Botnet. Botnet merupakan sekumpulan program yang saling terhubung melalui jaringan internet untuk melakukan tugas tertentu. Botnet merupakan program yang telah terinfeksi malware dan berada di bawah kendali aktor jahat. Botnet yang telah masuk ke dalam jaringan IoT akan menyebabkan penggunanya merasa dirugikan. Untuk menghadapi tantangan tersebut, digunakan kombinasi algoritma yaitu Decision Tree Classifier, Random Forest Classifier, dan Explainable Artificial Intelligence (XAI). Klasifikasi dataset N-BaIoT akan dilakukan untuk mempelajari pola dari botnet. Sebelum menghadapi proses klasifikasi, dataset telah dipersiapkan pada tahap preprocessing, yaitu penanganan missing value, penanganan duplicate value, dan pengubahan tipe data menjadi integer. Klasifikasi dilakukan dengan menggunakan dua buah algoritma. Kinerja kedua algoritma ini akan diukur dengan menggunakan skor F1. Hasil klasifikasi akan diteliti untuk mengetahui fitur interaksi pada dataset menggunakan Explainable Artificial Intelligence (XAI). XAI merupakan sekumpulan proses yang memungkinkan manusia untuk memahami hasil yang dihasilkan oleh algoritma machine learning. Penelitian ini menggunakan XAI untuk memahami hasil yang dihasilkan oleh Decision Tree Classifier dan Random Forest Classifier. Dengan demikian, penelitian ini menerapkan gabungan dari algoritma tersebut untuk mengetahui aktivitas fitur pada hasil keputusan algoritma tersebut
=====================================================================================================================================
The Internet of Things, which is usually called IoT, is one of the proofs of technological advancement. IoT provides its users with a lot of benefits and offers help in everyday life. However, not only does IoT offer benefits, it also brings challenges to its users. The challenge is that there is suspicious activity in the IoT traffic network. This suspicious activity is often referred to as a botnet, IoT Botnet. A botnet is a group of programs that are connected to each other via the internet network to perform certain tasks. Botnet refers to a program that has been infected with malware and is under the control of a malicious actor. Botnets that have entered the IoT network will cause users to feel disadvantaged. To face these challenges, the combination of algorithms used are Decision Tree Classifier, Random Forest Classifier, and Explainable Artificial Intelligence (XAI). Classification of the N-BaIoT dataset will be performed to learn the patterns of botnets. Before facing the classification process, the dataset has been prepared at the preprocessing stage, namely handling missing values, handling duplicate values, and changing the data type to integer. Classification is carried out using two algorithms. The performance of these two algorithms will be measured using the F1 score. The results of the classification will be investigated to find out the interaction features in the dataset using Explainable Artificial Intelligence (XAI). XAI is a set of processes that allow humans to understand the results created by machine learning algorithms. This research uses XAI to understand the results made by Decision Tree Classifier and Random Forest Classifier. Thus, this study applies a combination of these algorithms to determine the activity of features in the algorithm's decision results.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Classification, IoT Botnet, Decision Tree, Random Forest, Explainable ArtificialIntelligence (XAI).
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Rahel Cecilia Purba
Date Deposited:	02 Aug 2024 07:55
Last Modified:	24 Sep 2024 03:57
URI:	http://repository.its.ac.id/id/eprint/111703

Actions (login required)

View Item