hemz19-05/Algerian-Forest-Fires-Prediction-App

# Algerian Forest Fire Prediction (ML + Flask + AWS Deployment) This project predicts the Fire Weather Index (FWI) for Algerian forest fire regions using a Machine Learning model. The primary objective of this project was to practice: 🌐 Building and deploying a full end-to-end ML application ⚙️ Developing REST APIs using Flask ☁️ Deploying production-ready apps on AWS Elastic Beanstalk 🧩 Understanding real-world MLOps deployment workflows This project is based on the ML deployment course by Krish Naik, and I extended it with my own deployment + pipeline setup. ## 🚀 Demo Live App on AWS Elastic Beanstalk: 👉 algerianforestfirespredicti… ## 🔍 Problem Overview Forest fires are a major environmental and safety issue in Algeria. This application predicts the FWI using several meteorological features: 1. Temperature 2. Relative Humidity 3. Wind Speed 4. Rain 5. FFMC 6. DMC 7. ISI 8. Region + Classes ## 🧠 Model Used Ridge Regression (Scikit-Learn) Features scaled using StandardScaler Model artifacts saved as: model/ridg.pkl model/scaler.pkl ## 🧰 Tech Stack ML & Backend: ✔ Python ✔ Flask (REST APIs) ✔ Scikit-learn ✔ Pandas, NumPy Deployment & DevOps: ✔ AWS Elastic Beanstalk ✔ EC2 ✔ Gunicorn ✔ .ebextensions config ✔ Linux virtual server environment ## ⚙️ Application Architecture app/ │ ├── application.py # Flask API ├── wsgi.py # Entry point for Gunicorn ├── model/ # ML artifacts ├── templates/ # HTML interface ├── requirements.txt ├── Procfile # Gunicorn entry └── .ebextensions/ # AWS config ## 🖥️ How It Works You enter the inputs: Temperature, Rain, RH, Wind, FFMC, DMC, ISI, etc. The app: ✔ Validates form input ✔ Passes it into the model ✔ Predicts the Fire Weather Index (FWI) ✔ Returns the result to the UI ## ☁️ Deployment Workflow 🔥 Full deployment done on AWS: ✔Packaged app into deployment zip ✔ …