sebiakods/algeria-forest-fire-prediction-

Forest fire prediction system for Algeria using Random Forest (96.7% accuracy). Real-time API, interactive maps, and multi-model comparison (Logistic Regression, SVM, Neural Networks). Built with Python, scikit-learn, Flask, and Folium. # Algeria Forest Fire Prediction System A machine learning system to predict forest fire risks in Algeria using meteorological data and geospatial analysis ## Project Overview This project uses Random Forest classification to predict forest fire occurrences in Algerian regions based on weather conditions and fire weather indices. ### Key Features: - **Fire Risk Prediction** with 85%+ accuracy - **Interactive Map** visualization of high-risk regions - **REST API** for real-time predictions - **Multiple ML models** comparison (Logistic Regression, Decision Tree, SVM, Neural Network) ## Dataset The dataset contains 244 instances from Algerian regions with features including: - Temperature, Relative Humidity (RH), Wind Speed (Ws), Rain - Fire Weather Index components: FFMC, DMC, DC, ISI, BUI, FWI - Regions: Bejaia, Bouira, Chlef, El Tarf, Guelma, Jijel, Skikda, Tipaza, Tizi Ouzou ## Installation 1. Clone the repository: ```bash git clone github.com cd algeria-forest-fire-prediction