Analysis of Autoencoder Compression Performance in Intrusion Detection System

I. Gede Agung Krisna Pamungkas; Tohari Ahmad; Royyana Muslim Ijtihadie

doi:10.18280/ijsse.120314

Analysis of Autoencoder Compression Performance in Intrusion Detection System

I. Gede Agung Krisna Pamungkas, Tohari Ahmad^*, Royyana Muslim Ijtihadie

^*Corresponding author for this work

Institut Teknologi Sepuluh Nopember

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

Abstract

Exchanging data between devices is getting easier and faster just by using a network. Nevertheless, many factors threaten this process and the network itself. Implementing an Intrusion Detection System (IDS) may minimize the risk since it can identify and prevent attacks on the network. There are many methods to design an IDS to work optimally only by reducing data dimensions, one of which is by using the Autoencoder. However, its data dimensions may not have been optimal, which affects the IDS performance. In this study, we work on this problem. This study shows that one of the dimensional reduction methods can get optimal results. It indicates that it is implementable to secure the network.

Original language	English
Pages (from-to)	395-401
Number of pages	7
Journal	International Journal of Safety and Security Engineering
Volume	12
Issue number	3
DOIs	https://doi.org/10.18280/ijsse.120314
Publication status	Published - Jun 2022

Keywords

computer network
intrusion detection system
network infrastructure
network security

Access to Document

10.18280/ijsse.120314

Cite this

@article{135240d57eeb4f0c946960cbf6eaf1b5,

title = "Analysis of Autoencoder Compression Performance in Intrusion Detection System",

abstract = "Exchanging data between devices is getting easier and faster just by using a network. Nevertheless, many factors threaten this process and the network itself. Implementing an Intrusion Detection System (IDS) may minimize the risk since it can identify and prevent attacks on the network. There are many methods to design an IDS to work optimally only by reducing data dimensions, one of which is by using the Autoencoder. However, its data dimensions may not have been optimal, which affects the IDS performance. In this study, we work on this problem. This study shows that one of the dimensional reduction methods can get optimal results. It indicates that it is implementable to secure the network.",

keywords = "computer network, intrusion detection system, network infrastructure, network security",

author = "Pamungkas, {I. Gede Agung Krisna} and Tohari Ahmad and Ijtihadie, {Royyana Muslim}",

year = "2022",

month = jun,

doi = "10.18280/ijsse.120314",

language = "English",

volume = "12",

pages = "395--401",

journal = "International Journal of Safety and Security Engineering",

issn = "2041-9031",

publisher = "International Information and Engineering Technology Association",

number = "3",

}

TY - JOUR

T1 - Analysis of Autoencoder Compression Performance in Intrusion Detection System

AU - Pamungkas, I. Gede Agung Krisna

AU - Ahmad, Tohari

AU - Ijtihadie, Royyana Muslim

PY - 2022/6

Y1 - 2022/6

N2 - Exchanging data between devices is getting easier and faster just by using a network. Nevertheless, many factors threaten this process and the network itself. Implementing an Intrusion Detection System (IDS) may minimize the risk since it can identify and prevent attacks on the network. There are many methods to design an IDS to work optimally only by reducing data dimensions, one of which is by using the Autoencoder. However, its data dimensions may not have been optimal, which affects the IDS performance. In this study, we work on this problem. This study shows that one of the dimensional reduction methods can get optimal results. It indicates that it is implementable to secure the network.

AB - Exchanging data between devices is getting easier and faster just by using a network. Nevertheless, many factors threaten this process and the network itself. Implementing an Intrusion Detection System (IDS) may minimize the risk since it can identify and prevent attacks on the network. There are many methods to design an IDS to work optimally only by reducing data dimensions, one of which is by using the Autoencoder. However, its data dimensions may not have been optimal, which affects the IDS performance. In this study, we work on this problem. This study shows that one of the dimensional reduction methods can get optimal results. It indicates that it is implementable to secure the network.

KW - computer network

KW - intrusion detection system

KW - network infrastructure

KW - network security

UR - http://www.scopus.com/inward/record.url?scp=85135580481&partnerID=8YFLogxK

U2 - 10.18280/ijsse.120314

DO - 10.18280/ijsse.120314

M3 - Article

AN - SCOPUS:85135580481

SN - 2041-9031

VL - 12

SP - 395

EP - 401

JO - International Journal of Safety and Security Engineering

JF - International Journal of Safety and Security Engineering

IS - 3

ER -

Analysis of Autoencoder Compression Performance in Intrusion Detection System

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this