mayank-kumar03/FWI_calculator_algerian_forest
Domaine:
environment and energyType de record:
software
Créateur:
may
Hôte:
# 🔥 Forest Fire Area Prediction
---
## 🌱 About the Project
**Forest Fire Area Prediction** is a machine learning web application that predicts the **area affected by a forest fire** based on several environmental and regional parameters.
### R2 Score : 0.9842993364555512
🔗 **Live App**: Click Here to Use the Application
## 🎯 Real-World Use Case
Forest fires are becoming more frequent due to climate change. Early prediction of the **scale of destruction** helps stakeholders take preventive actions. This system:
- 📈 Predicts **how much area might burn** under specific weather conditions.
- ⚠️ Aims to **assist forest departments, environmental agencies**, and **disaster response units**.
- 🌱 Helps preserve **biodiversity**, **forests**, and **minimize CO₂ emissions**.
## 🖼 Screenshots
---
## 🚀 Tech Stack
- 🐍 Python 3.8+
- 🌐 Flask (Web Framework)
- 🤖 Scikit-learn (Machine Learning)
- 📦 Pickle for model storage
- 🧪 Ridge Regression
- 🖥️ HTML (Jinja templates)
---
## 🧠 How It Works
1. User inputs weather and fire parameters in the web form.
2. Data is normalized using `StandardScaler`.
3. A trained Ridge Regression model predicts the affected area.
4. The result is displayed on a separate webpage.
---
## 🔍 Interpreting FWI Values
| **FWI Value** | **Fire Danger Level** | **Meaning** |
|-----------------|------------------------|-------------|
| 0.0 – 5.0 | 🔵 **Low** | Fires are unlikely or easily controlled. |
| 5.1 – 12.0 | 🟡 **Moderate** | Fires may start and require attention. |
| 12.1 – 30.0 | 🟠 **High** | Fires spread quickly and require active suppression. |
| 30.1 – 50.0 | 🔴 **Very High** | Fires ignite easily, spread rapidly. |
| > 50.0 | ⚫ **Extreme** | Explosive fire behavior; immediate emergency action needed. |
---
## 📥 Input Parameters
| Feature | Description |
|------------|------------------------------------|
| Temperature| …