🛡️ Network Intrusion Detection System using Machine Learning & Deep Learning

A Network Intrusion Detection System (NIDS) built using Machine Learning and Deep Learning techniques on the NSL-KDD dataset. This project has been modernized to run seamlessly on Google Colab, Python 3.12, and TensorFlow/Keras 3.

---

📌 Project Overview

Network Intrusion Detection Systems are essential in cybersecurity for identifying malicious network traffic and protecting systems from unauthorized access.

This project implements an end-to-end intrusion detection pipeline that preprocesses network traffic data, trains multiple Machine Learning models, builds an LSTM-based Deep Learning model, and evaluates their performance using standard classification metrics.

---

✨ Features

• Data preprocessing and feature engineering
• Binary and multi-class intrusion detection
• Multiple Machine Learning classifiers
• LSTM-based Deep Learning model
• Model performance comparison
• ROC Curve and training visualization
• Google Colab compatible
• Python 3.12 and TensorFlow/Keras 3 compatible

---

🛠️ Technologies Used

• Python
• TensorFlow / Keras
• Scikit-learn
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Google Colab

---

📂 Repository Structure

Network-Intrusion-Detection-using-MachineLearning/
│
├── README.md
├── documentation.md
├── LICENSE
├── requirements.txt
├── .gitignore
│
├── Data_Preprocessing_NSL_KDD.ipynb
├── Classifiers_NSL_KDD.ipynb
├── Intrusion_Detection_NSL_KDD.ipynb
│
├── datasets/
│   └── README.md
│
├── images/
│   ├── workflow.png
│   ├── model_architecture.png
│   ├── training_accuracy.png
│   ├── training_loss.png
│   └── roc_curve.png
│
├── labels/
└── plots/

---

📖 Project Workflow

The workflow begins with the NSL-KDD dataset, followed by data preprocessing, feature engineering, training multiple Machine Learning models, building an LSTM-based Deep Learning model, evaluating performance, and finally predicting network intrusions.

---

🧠 Model Architecture

The Deep Learning model is based on an LSTM network followed by a Dense output layer for intrusion classification.

---

📊 Model Performance

Training Accuracy

Training Loss

ROC Curve

---

📊 Dataset

This project uses the NSL-KDD dataset, a benchmark dataset widely used for evaluating Network Intrusion Detection Systems.

The dataset is not included in this repository due to its size.

Please download it separately and place it inside the datasets/ folder before running the notebooks.

---

🚀 Getting Started

Clone the Repository

git clone https://github.com/parishree-gupta/Network-Intrusion-Detection-using-MachineLearning.git

Install Dependencies

pip install -r requirements.txt

Run the Notebooks

Execute the notebooks in the following order:

1. Data_Preprocessing_NSL_KDD.ipynb
2. Classifiers_NSL_KDD.ipynb
3. Intrusion_Detection_NSL_KDD.ipynb

---

📈 Results

The implemented models are evaluated using multiple performance metrics, including:

• Accuracy
• Precision
• Recall
• F1-Score
• ROC Curve
• Confusion Matrix

These metrics help assess the effectiveness of the intrusion detection models in distinguishing normal and malicious network traffic.

---

🔄 Modernization

This repository has been updated from the original implementation to support the latest software ecosystem.

Improvements

• ✅ Google Colab compatible
• ✅ Python 3.12 support
• ✅ TensorFlow/Keras 3 compatibility
• ✅ NumPy 2.x compatibility
• ✅ pandas 2.x compatibility
• ✅ Updated deprecated APIs
• ✅ Improved notebook execution and stability

---

💡 Future Enhancements

• Real-time packet capture
• Live intrusion detection dashboard
• Explainable AI (XAI)
• Docker deployment
• Cloud deployment
• REST API integration

---

👩‍💻 Author

Pari Shree Gupta

B.Tech Computer Science Engineering

Jaypee Institute of Information Technology

GitHub: https://github.com/parishree-gupta

---

📜 License

This project is licensed under the MIT License.

---

⭐ Support

If you found this project useful, consider giving it a ⭐ on GitHub!