🌲 Algerian Forest Fire Prediction – ML Web App
A production-ready Machine Learning web application that predicts the Fire Weather Index (FWI) using meteorological and fire weather index parameters from Algerian forest data.
The application is built using Flask and deployed on Render for scalable cloud-based inference.
🚀 Live Application
Live URL:
algerian-forest-1.onrender.…
🧠 Problem Statement
Forest fires cause severe environmental and economic damage. Early prediction of fire risk helps authorities take preventive measures and manage resources effectively.
This project predicts the Fire Weather Index (FWI) using meteorological conditions and fire weather indices from Algerian forest regions. The system allows users to input environmental parameters and instantly receive predictions through a web interface.
📊 Dataset
Dataset Name: Algerian Forest Fires Dataset
Regions Covered:
Bejaia
Sidi Bel-Abbes
The dataset contains meteorological observations and fire weather indices collected from Algerian forests.
🔢 Input Features
Temperature
Relative Humidity (RH)
Wind Speed (Ws)
Rain
FFMC (Fine Fuel Moisture Code)
DMC (Duff Moisture Code)
ISI (Initial Spread Index)
Classes (Fire / No Fire)
Region (Bejaia / Sidi Bel-Abbes)
🛠️ Tech Stack
Machine Learning
Python 3.10
NumPy
Pandas
Scikit-learn
Ridge Regression
Web Development
Flask
HTML
CSS
Deployment
Gunicorn
Render
☁️ Cloud Deployment
Application deployed on Render as a Web Service
Gunicorn used as production WSGI server
Python runtime environment used for hosting
Real-time prediction system for machine learning inference
📈 Model Details
Algorithm: Ridge Regression
Preprocessing: Standard Scaling
Python Version: 3.10
Model Persistence: Pickle (.pkl files)
✅ Key Learnings
End-to-end machine learning project workflow
Data preprocessing and feature engineering
Model training and evaluation
Model serialization usi …