This project explores machine learning techniques for network intrusion detection using the UNSW-NB15 cybersecurity dataset. The goal of the project is to analyze network traffic data and build predictive models capable of identifying malicious activity.
The analysis focuses on data preprocessing, feature analysis, and machine learning classification models implemented in R.
The project uses the UNSW-NB15 dataset, a modern network intrusion detection dataset created by the Australian Centre for Cyber Security.
The dataset contains network traffic features used to classify whether traffic is normal or malicious.
- Clean and preprocess the dataset
- Perform exploratory data analysis
- Identify important features
- Train machine learning models
- Evaluate model performance
The following models were implemented:
- Logistic Regression
- Random Forest
These models were used to classify network traffic as normal or intrusion.
- R
- Data Cleaning & Preprocessing
- Logistic Regression
- Random Forest
- Data Visualization
unsw-nb15-intrusion-detection │ ├── analysis.R # Main analysis script ├── dataset.csv # Sample dataset (optional) ├── plots/ # Visualizations and graphs └── README.md
- Load and preprocess the dataset
- Convert required variables to numeric
- Perform exploratory data analysis
- Train machine learning models
- Evaluate model performance
Krish Sharma
BSc Computer Science (Artificial Intelligence)
Asia Pacific University of Technology & Innovation
GitHub:
https://github.com/krishsharma5169
LinkedIn:
https://www.linkedin.com/in/krish-sharma-2457a322a/
This project was developed as part of coursework for Programming for Data Analysis and demonstrates the application of machine learning techniques for cybersecurity problems.