deveshmaithani/Algerian-Forest-Fire-Prediction

This project predicts forest fires in Algeria using the Algerian Forest Fires Dataset. # Algerian-Forest-Fire-Prediction This project predicts forest fires in Algeria using the Algerian Forest Fires Dataset. It includes Exploratory Data Analysis (EDA), Feature Engineering, model training, and a Flask app for predictions. ## Data Preprocessing & Feature Engineering ### 1. Exploratory Data Analysis (EDA) Understanding feature distributions Detecting missing values and outliers Correlation analysis between features ### 2. Feature Engineering Handling missing values Encoding categorical features Scaling numerical variables ## Models Used The project evaluates the following regression models: Linear Regression Lasso Regression Ridge Regression ElasticNet Regression #### After comparison, the best-performing model is selected for deployment. ## Tech Stack Python (NumPy, Pandas, Scikit-learn, Matplotlib, Seaborn) Machine Learning (Regression Models) Flask (Web Application for Predictions) HTML/CSS (Frontend for Flask App) ## Project Workflow 1. EDA & Feature Engineering 2. Model Training & Evaluation 3. Flask App Development 4. Deployment (Optional) ## Project Structure /Algerian_Forest_Fires_Prediction/ │── application.py # Flask app for user input and predictions │── scaler.pkl/ ridge.pkl # Trained ML model (Pickle file) │── static/ # CSS, images (if any) │── templates/ │ ├── index.html # Frontend UI │── data/ │ ├── Algerian_forest_fires_dataset.csv # Dataset │── notebooks/ # Jupyter Notebooks for EDA & Model Training │── scripts/ # Python scripts for preprocessing & training │── README.md # Documentation │── requirements.txt # Dependencies ## Model Performance Model RMSE MAE R2 Score Linear Regression 0.5468 0.9847 Lasso Regression 1.1331 0.949 Ridge Regression 0.564 0.564 ElasticNet Regression 1.882 0.8753 ## Installation & Usage 1 Clone the Repository git clone github.com …