title	Forest Fire Prediction
emoji	📚
colorFrom	indigo
colorTo	gray
sdk	gradio
sdk_version	5.25.2
app_file	app.py
pinned	false
license	mit
short_description	Predict forest fire risk using Fire Weather Index (FWI)

Forest Fire Prediction

About

A Gradio regression application that estimates the Fire Weather Index from environmental inputs and maps the result to fire-risk guidance. The repository includes a Ridge-based model workflow and a Hugging Face Spaces configuration; evaluation is dataset- and split-specific.

---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

📋 Table of Contents

• Overview
• Features
• Demo
• Installation
• Usage
• Input Parameters
• Risk Levels
• Technical Details
• Project Structure
• Contributing
• License

🌟 Overview

This machine learning-based system predicts the Fire Weather Index (FWI) in forest areas using meteorological and environmental factors. It provides real-time risk assessment and recommendations for fire prevention and management.

✨ Features

• Real-time Predictions: Instant FWI calculations
• Interactive Interface: User-friendly web interface with intuitive controls
• Risk Assessment: Three-level risk classification system
• Smart Recommendations: Contextual advice based on risk levels
• Input Validation: Comprehensive parameter validation
• Visual Feedback: Color-coded results and clear visual indicators

🎮 Demo

Try the live demo on Hugging Face Spaces: Forest Fire Prediction Demo

🚀 Installation

Prerequisites

• Python 3.8 or higher
• pip package manager

Setup Steps

# Clone the repository
git clone https://github.com/dhananjaynerkar/forest-fire-prediction.git
cd forest-fire-prediction

# Create and activate virtual environment
python -m venv venv
source venv/bin/activate  # Linux/Mac
venv\Scripts\activate     # Windows

# Install dependencies
pip install -r requirements.txt

# Run the application
python app.py

💻 Usage

1. Access the web interface at http://localhost:7861
2. Input environmental parameters using the sliders
3. Select the region and fire class
4. Click "Predict Fire Weather Index"
5. Review the prediction, risk level, and recommendations

📊 Input Parameters

Parameter	Range	Description
Temperature	22-42°C	Ambient temperature
Relative Humidity	21-90%	Air humidity percentage
Wind Speed	6-29 km/h	Wind velocity
Rain	0-16.8 mm	Rainfall amount
FFMC	28.6-92.5	Fine Fuel Moisture Code
DMC	1.1-65.9	Duff Moisture Code
ISI	0-18.5	Initial Spread Index
Classes	0 or 1	Fire occurrence (0: No, 1: Yes)
Region	0 or 1	Location (0: Bejaia, 1: Sidi-Bel Abbes)

🎯 Risk Levels

Low Risk (0-10)

• Regular monitoring recommended
• Standard fire prevention measures
• Suitable for controlled burns

Moderate Risk (11-20)

• Enhanced monitoring required
• Increased vigilance needed
• Review of fire response procedures

High Risk (>20)

• Immediate precautions necessary
• Constant monitoring required
• Emergency response team standby

🔧 Technical Details

Model Architecture

• Algorithm: Ridge Regression with Cross-Validation (RidgeCV)

  • Selected for its robust performance with continuous target variables
  • Handles multicollinearity in weather-related features
  • Reported checkpoint: R² 0.9843 and MAE 0.5642 on a 61-row holdout; this small-sample result is not a production generalization guarantee.

• Feature Engineering:

  • Input features correlation analysis (threshold: 0.85)
  • Removed highly correlated features to prevent multicollinearity
  • Feature scaling using StandardScaler
  • Normalized numerical features to [0,1] range

• Model Pipeline:

  1. Input Validation: Range-based checks for all parameters
  2. Feature Preprocessing: StandardScaler transformation
  3. Prediction: Ridge Regression model
  4. Post-processing: Risk level classification and recommendations

• Performance Metrics:

  • Cross-validation: 5-fold CV
  • Evaluation metrics: MAE, R² Score
  • Model persistence using pickle serialization

Technologies Used

• Backend: Python, scikit-learn
• Frontend: Gradio 5.25.2+
• Data Processing: NumPy, Pandas
• Model Serialization: Pickle
• Deployment: Hugging Face Spaces
• Version Control: Git LFS for model files

📁 Project Structure

forest-fire-prediction/
│
├── app.py                 # Main application file
├── dataset/              # Dataset directory
│   └── Algerian_forest_fires_cleaned_dataset.csv
├── models/               # Model directory
│   ├── ridge.pkl         # Trained model
│   └── scaler.pkl        # Fitted scaler
├── notebooks/            # Jupyter notebooks
│   ├── EDA And FE Algerian Forest Fires.ipynb
│   └── Model Training.ipynb
├── venv/                 # Virtual environment
├── .env                  # Environment variables
├── .env.example          # Environment variables template
├── .gitattributes        # Git LFS configuration
├── .gitignore           # Git ignore rules
├── demo1.PNG             # Demo screenshots
├── demo2.PNG
├── demo3.PNG
├── deploy.py             # Deployment script
└── requirements.txt     # Dependencies

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit changes (git commit -m 'Add AmazingFeature')
4. Push to branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

Distributed under the MIT License. See LICENSE for more information.

👥 Authors

• Dhananjay Nerkar - GitHub

🙏 Acknowledgments

• UCI Machine Learning Repository for providing the Algerian Forest Fires Dataset
• The research paper: "Predicting Forest Fire in Algeria using Data Mining Techniques: Case Study of the Decision Tree Algorithm" by Faroudja ABID et al.
• Hugging Face for hosting the demo application
• Gradio team for the excellent UI framework
• All contributors and reviewers who helped improve this project

📧 Contact

Dhananjay Nerkar - nerkarr.dhananjay@gmail.com Project Link: https://github.com/dhananjaynerkar/forest-fire-prediction