Anomaly-Based Network Intrusion Detection System

Authors

  • Anil Verma Department of Electronics and Computer Engineering, Pulchowk Campus, Tribhuvan University, Pulchowk, Lalitpur, Nepal
  • Enish Paneru Department of Electronics and Computer Engineering, Pulchowk Campus, Tribhuvan University, Pulchowk, Lalitpur, Nepal
  • Bishal Baaniya Department of Electronics and Computer Engineering, Pulchowk Campus, Tribhuvan University, Pulchowk, Lalitpur, Nepal

DOI:

https://doi.org/10.3126/lecj.v4i1.49364

Keywords:

Network, Intrusion, Artificial Intelligence, Random Forest, KNN

Abstract

Network security has been a really hot topic since the inception of the internet in the early ’80s. With millions of people entrusting their life savings in the hands of an organization, it is really necessary to keep the network intruders out of the system. The most alarming thing is that - even today, many organizations are detecting these intrusions through manual labour. Many researchers have proven that these intrusions have a certain pattern i.e. they can be detected with an Artificial Intelligence (AI) based system with enough training which can prove to be a really an effective substitute for manual labour. This paper explains the current trends in Network Intrusion Detection and the technologies that have been implemented to detect them. CICIDS2017 dataset containing around 3 million data points was used in this experiment. K-Nearest Neighbours (KNN) and Random Forest algorithms are used as the AI tools and their performance has also been compared.

Downloads

Download data is not yet available.
Abstract
21
PDF
37

Downloads

Published

2022-12-07

How to Cite

Verma, A., Paneru, E. and Baaniya, B. (2022) “Anomaly-Based Network Intrusion Detection System”, Journal of Lumbini Engineering College, 4(1), pp. 38–42. doi: 10.3126/lecj.v4i1.49364.

Issue

Section

Research Articles