DDoS Protection Model

Welcome to the AI DDoS Protection System! This project leverages machine learning to identify and mitigate potential DDoS attacks using various algorithms. The application features a user-friendly interface to input network parameters and predict the likelihood of a DDoS attack.

Live Demo

You can see the live version of the application at this URL.

Table of Contents

• Features
• Installation
• Usage
• Model Performance
• Technologies Used
• Dataset
• Contributing
• License

Features

• Real-time prediction of DDoS attacks.
• User-friendly interface for inputting network parameters.
• Visual representation of model performance metrics.

Installation

To get started, clone this repository and install the required packages:

git clone https://github.com/nky001/ddos.git
cd your_repository
pip install -r requirements.txt

Requirements

Make sure you have the following Python packages installed:

• Flask
• scikit-learn
• joblib
• flask_socketio
• pandas
• numpy
• eventlet

You can create a requirements.txt file with the following content:

Flask
scikit-learn
joblib
flask_socketio
pandas
numpy
eventlet

Usage

Run the application using Gunicorn:

gunicorn -k gevent -w 1 app:app

Open your web browser and navigate to http://localhost:5000 to access the app.

Input Parameters

You can input the following network parameters to predict the likelihood of a DDoS attack:

• Flow Duration
• Source IP
• Source Port
• Destination IP
• Destination Port
• Total Forward Packets
• Initial Backward Window Bytes
• Protocol

Model Performance

Gaussian Naive Bayes

• Training-set accuracy score: 81.90%
• Model accuracy score: 81.87%

Random Forest Classifier

• Cross-validation accuracy scores for each fold: [1.0, 1.0, 0.99999844, 0.99999844, 1.0]
• Mean cross-validation accuracy: 99.9999%
• Standard deviation: 7.66e-07

The models were trained using various machine learning techniques to ensure high accuracy in DDoS detection.

Dataset

The project uses a DDoS dataset sourced from Kaggle. You can download it using the following command:

import kagglehub

path = kagglehub.dataset_download("devendra416/ddos-datasets")
print("Path to dataset files:", path)

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Feel free to reach out if you have any questions or need further assistance. Thank you for checking out this.