Comparative Analysis of Deep Learning Models and Preprocessing Techniques for Anomaly Detection in Syslog

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21260%2F23%3A00370775" target="_blank" >RIV/68407700:21260/23:00370775 - isvavai.cz</a>

Result on the web

<a href="https://doi.org/10.1109/ICICS60529.2023.10330520" target="_blank" >https://doi.org/10.1109/ICICS60529.2023.10330520</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICICS60529.2023.10330520" target="_blank" >10.1109/ICICS60529.2023.10330520</a>

Result language

angličtina

Original language name

Comparative Analysis of Deep Learning Models and Preprocessing Techniques for Anomaly Detection in Syslog

Original language description

With the increasing number of cybersecurity attacks and their increasing complexity, it is necessary to adapt the detection methods to be able to prevent such attacks. The replacement of traditional detection methods based on machine learning with more advanced deep learning and neural network approaches is the crucial step for this. In this paper, we present a comparative analysis of different deep learning models for anomaly detection in syslog. We analysed existing datasets for system logs and compared several preprocessing methods. We evaluated different deep learning models on those preprocessed datasets to provide a comprehensive overview of the current state-of-the-art in cybersecurity. We achieved the best results with the CNN model with 0.999 F1-score on the BGL dataset showing the great potential in such techniques for the real-life models monitoring the system and detecting anomalies.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

<a href="/en/project/TM03000055" target="_blank" >TM03000055: Multidimensional detection and automated response using artificial intelligence</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2023

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

2023 14th International Conference on Information and Communication Systems (ICICS)

ISBN

979-8-3503-0786-3

ISSN

2471-125X

e-ISSN

2573-3346

Number of pages

6

Pages from-to

1-6

Publisher name

IEEE Jordan Section

Place of publication

—

Event location

Irbid

Event date

Nov 21, 2023

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—