Network Intrusion Detection using NSL_KDD Dataset
Partially based on the research paper https://www.sciencedirect.com/science/article/pii/S0925231219315759?via%3Dihub
Network Intrusion Detection system for cyber threats, to enhance cybersecurity measures. The cybersecurity ML project aims to leverage machine learning algorithms to improve the security of computer networks and systems against cyber attacks. The project involves developing an intrusion detection system that can automatically identify and classify malicious activities in real time based on patterns and anomalies in network traffic.
Our first step was to make the best use of the available data, i.e. NSL-KDD, and extract the features to have maximum utilization. We then tried various models suitable for this, highlights being, multi layer peerceptron and LSTM
The experimental results show the need for proper manipulation of these features to reduce the overhead in modelling the IDS. scores. We got recall score of 0.9904 through LSTM .
The notebook can be run on
- Google Colaboratory
- Jupyter Notebook