Intrusion Detection System (IDS) - UNSW-NB15

This project implements a machine learning-based Intrusion Detection System (IDS) using the UNSW-NB15 dataset. It preprocesses the data, selects key features, trains multiple classification models, and saves the best-performing model for future use.

Features

• Data preprocessing and label encoding for categorical features
• Feature selection using F-score
• Training and evaluation of Random Forest, Gradient Boosting, and KNN classifiers
• Model comparison with accuracy and cross-validation
• Visualization of model performance
• Saving trained models and encoders for deployment

Dataset

• UNSW_NB15_training.csv: Training data
• UNSW_NB15_testing.csv: Testing data

Requirements

• Python 3.7+
• pandas
• scikit-learn
• seaborn
• matplotlib

Install dependencies with:

pip install pandas scikit-learn seaborn matplotlib

Usage

1. Prepare the data
   Place UNSW_NB15_training.csv and UNSW_NB15_testing.csv in the project directory.

2. Run the notebook
   Open reference_book_backend.ipynb in VS Code or Jupyter Notebook and run all cells.

3. Model Training
   The notebook will:

   • Preprocess and encode the data
   • Select the top 15 features
   • Train Random Forest, Gradient Boosting, and KNN models
   • Evaluate and compare their performance

4. Model Saving
   The best-performing model (Random Forest) and label encoders are saved as .pkl files for later use:

   • rf_model.pkl
   • le_service.pkl
   • le_state.pkl
   • le_protocol.pkl
   • le_attack.pkl

Project Structure

intrusion detection/
├── reference_book_backend.ipynb
├── UNSW_NB15_training.csv
├── UNSW_NB15_testing.csv
├── rf_model.pkl
├── le_service.pkl
├── le_state.pkl
├── le_protocol.pkl
├── le_attack.pkl
└── README.md

Results

• The Random Forest model achieved the highest accuracy among the tested models.
• Model performance is visualized using a bar plot.