Ravipandey16/Algerian-Forest-Fires-Dataset

This project uses the Algerian Forest Fires Dataset from the UCI Machine Learning Repository to build a Machine Learning model that predicts forest fire occurrence based on meteorological and Fire Weather Index (FWI) parameters. Algerian Forest Fires Dataset – ML Prediction Project This project uses the Algerian Forest Fires Dataset from the UCI Machine Learning Repository to build a Machine Learning model that predicts forest fire occurrence based on meteorological and Fire Weather Index (FWI) parameters. Dataset Source The dataset is publicly available on the UCI ML Repository: Algerian Forest Fires Dataset (UCI) Origin: 2012–2015 forest fire records collected from north-eastern Algeria Regions: Bejaia and Sidi Bel-Abbes Dataset Description The dataset contains 244 instances and 11 meteorological + FWI features, including: Temperature Relative Humidity (RH) Wind Speed (Ws) Rain Fine Fuel Moisture Code (FFMC) Duff Moisture Code (DMC) Initial Spread Index (ISI) Classes (Fire / No Fire) Region (Bejaia = 1, Sidi Bel-Abbes = 0) Project Objectives Clean and preprocess the Algerian forest fire dataset Perform EDA (Exploratory Data Analysis) Train and evaluate ML models for fire prediction Export the final model using Pickle Build a Flask web application for real-time prediction Deploy UI form for user input (Temperature, RH, Ws, FFMC, etc.) Machine Learning Workflow Data Preprocessing Feature Scaling Train/Test Split ML Algorithms Hyperparameter Tuning Model Evaluation (R², Accuracy, RMSE, etc.) Model Serialization (model.pkl) Flask Web App Features Clean HTML form for input Backend ML inference POST data handling Predictive output returned to UI Technologies Used Python pandas, numpy scikit-learn matplotlib, seaborn Flask HTML, CSS & JavaScript Pickle (Model Serialization) 📢 Note This project is for research and educational purposes using publicly available data from the UCI Machine Learning Repository.