kondalokesh025-code/Algerian_Forest_Fires_Prediction

Predicting the Fire Weather Index (FWI) for Algerian forest fires using regression models, comparing a few and picking the best one with RidgeCV. Built with Streamlit. # Algerian Forest Fire FWI Prediction A small project where I try to predict the Fire Weather Index (FWI) using the Algerian Forest Fires dataset. Rebuilt it with Streamlit this time instead of Flask. ## What this is about Instead of just picking one model and going with it, I tried a few and compared them: * Linear Regression * Ridge Regression * Lasso Regression * ElasticNet I also ran RidgeCV to do cross-validation, and it came out on top. So that's the one running the actual app now. ## What I did * Explored the data (EDA) * Cleaned it up * Looked at correlations and features * Built pipelines for the models * Tried out L1 and L2 regularization * Compared the models against each other * Cross-validated everything * Checked the results properly * Put it all in a Streamlit app ## About the dataset It's the Algerian Forest Fires dataset, weather and fire data from two regions in Algeria, Bejaia and Sidi Bel-Abbes. The thing I'm predicting is FWI (Fire Weather Index). ## Tools I used Python, Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn, Streamlit ## How to run it ```bash pip install -r requirements.txt streamlit run app.py ``` That's pretty much it. Check out the notebook if you want to see how I compared the models.