OPTIMALISASI KLASIFIKASI MULTICLASS SERANGAN SIBER DENGAN ALGORITMA PROXIMAL POLICY OPTIMIZATION (PPO) DAN ADVANTAGE ACTOR-CRITIC (A2C) PADA REINFORCEMENT LEARNING

TOYYIB, AHMED ATHALLAH and Heryanto, Ahmad (2025) OPTIMALISASI KLASIFIKASI MULTICLASS SERANGAN SIBER DENGAN ALGORITMA PROXIMAL POLICY OPTIMIZATION (PPO) DAN ADVANTAGE ACTOR-CRITIC (A2C) PADA REINFORCEMENT LEARNING. Undergraduate thesis, Sriwijaya University.

Preview	Image RAMA_56201_09011382126165_cover.png - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (332kB) | Preview
	Text RAMA_56201_09011382126165.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (4MB) | Request a copy
	Text RAMA_56201_09011382126165_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (14MB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_56201_09011382126165_0022018703_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (335kB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (857kB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (15kB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (99kB) | Request a copy
	Text RAMA_56201_09011382126165_0022018703_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (893kB) | Request a copy

Abstract

The Proximal Policy Optimization (PPO) and Advantage Actor-Critic (A2C) methods have proven effective in detecting, evaluating, and performing multi-class classification of cyberattacks. This study implements both algorithms within a Reinforcement Learning framework to classify types of attacks based on network traffic features. The datasets used for testing include CIC-IDS2018, CIC-IDS2017, ISCX2012, and NSL-KDD. Each dataset undergoes preprocessing, normalization, and feature selection using the SelectKBest method to obtain the most relevant features. Experimental results show that both PPO and A2C algorithms are capable of detecting attacks with high accuracy, with performance variations depending on the characteristics of the dataset. The PPO method excels in training stability and reward utilization, while A2C demonstrates strong adaptability to continuous exploitation strategies. With a careful approach to feature selection, data ratio, and model parameters, this system can deliver accurate and efficient detection in modern multi-class cyberattack classification

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Klasifikasi Serangan Siber, Reinforcement Learning, Proximal Policy Optimization (PPO), Advantage Actor-Critic (A2C)
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1-2040 Engineering (General). Civil engineering (General) > TA158.7 Computer network resources Including the Internet Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4150-4380 Computer network resources
Divisions:	09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User:	Ahmed Athallah Toyyib
Date Deposited:	11 Jun 2025 01:54
Last Modified:	11 Jun 2025 01:54
URI:	http://repository.unsri.ac.id/id/eprint/175324

Actions (login required)

View Item