AsmitaKabra/Algerian_forest_fires_model

End-to-end ML application for predicting forest fire risk using the Algerian Forest Fires dataset. # Forest Fire Prediction - End-to-End ML Web Application ## Project Overview This project is an end-to-end Machine Learning web application developed to predict forest fire risk using environmental and weather-related parameters. The application allows users to input values such as temperature, humidity, wind speed, rain, and other fire weather indices to generate predictions using a trained Ridge Regression model. The project demonstrates: - Data preprocessing - Feature engineering - Exploratory Data Analysis (EDA) - Model training and evaluation - Model deployment using Flask - Frontend integration using HTML/CSS --- ## Tech Stack - Python - Flask - Scikit-learn - Pandas - NumPy - HTML/CSS - Pickle --- ## Features - User-friendly web interface - Real-time prediction generation - Data scaling using StandardScaler - Ridge Regression model integration - End-to-end ML pipeline --- ## Project Structure ```bash project/ │ ├── application.py ├── requirements.txt ├── README.md │ ├── models/ │ ├── ridge.pkl │ └── scaler.pkl │ ├── templates/ │ ├── index.html │ └── home.html │ ├── notebooks/ │ ├── EDA&FeatureSelection.ipynb │ └── model training.ipynb │ ├── screenshots/ │ ├── home.png │ └── result.png ``` --- ## Machine Learning Workflow 1. Data Collection 2. Data Cleaning 3. Exploratory Data Analysis 4. Feature Selection 5. Data Scaling 6. Model Training 7. Model Evaluation 8. Model Serialization using Pickle 9. Flask App Development 10. Deployment Ready Application --- ## Input Parameters The model uses the following features: - Temperature - Relative Humidity (RH) - Wind Speed (Ws) - Rain - FFMC - DMC - ISI - Classes - Region --- ## Model Used ### Ridge Regression Reason for selection: - Helps reduce overfitting - Performs well on multicollinearity - Provides stable predictions --- ## Installation & Setup ### Clone Re …