Predicts Fire Weather Index using machine learning on Algerian Forest Fires dataset. Six regression models tested: Linear, Ridge, Ridge CV, Lasso, Lasso CV, Elastic Net. Best Model: Ridge CV with R² 0.984, MAE 0.564, RMSE 0.834. Features: Flask web app, 7-tab interactive dashboard, real-time FWI predictions
# 🔥 Forest Fire Prediction - Machine Learning Regression Analysis
## 📋 Table of Contents
- Overview
- Dataset
- Models & Results
- Mathematical Framework
- Technical Stack
- Installation & Setup
- Usage
- Project Structure
- Features
- Visualizations
- Results & Performance
- Key Insights
- Future Improvements
- References
---
## 📖 Overview
This project implements a comprehensive end-to-end machine learning solution for predicting the **Fire Weather Index (FWI)** using advanced regression models. The system analyzes meteorological data and Fire Weather Index system components from the **Algerian Forest Fires dataset (2012)** to build predictive models for forest fire risk assessment.
The project includes:
- **6 Regression Models** trained and evaluated
- **Interactive Flask Web Application** with real-time predictions
- **Comprehensive ML Dashboard** with 7 feature-rich tabs
- **Mathematical Framework** with detailed equations
- **9 High-Quality Visualizations** (300 DPI)
- **Responsive Light Theme** with modern design
- **Production-Ready Code** with proper structure
**Live Demo**: Access the dashboard at `
localhost` after running the application.
---
## 📊 Dataset
### Source Information
- **Dataset**: Algerian Forest Fires (2012)
- **Regions**: Bejaia and Sidi-Bel Abbes (Algeria)
- **Time Period**: June - September 2012
- **Total Instances**: 244 fire records
- **Features**: 11 input variables
### Features Description
| Feature | Type | Range | Unit |
|---------|------|-------|------|
| Temperature | Numeric | 22 - 42 | °C |
| Relative Humidity | Numeric | 21 - 90 | % |
| Wind Speed | Numeric | 6 - 29 | km/h |
| Rainfall | Numeric | 0 - 16.8 | mm |
| FFMC Index | Numeric | 28.6 - 92.5 | - |
| DMC Index | Numeric | 1.1 - 65.9 | - |
| ISI Index | Numeric | 0 - 18.5 | - |
| Region | Categorical | 0, 1 | - |
| Class | Categorical | 0, 1 | - |
### FWI System Components
The **Fire Weather Index (FWI)** system consists of three moisture codes …