Comparative Evaluation of Random Forest, XGBoost, and SVC Models for Fire Detection

Both natural and man-made ecosystems continue to face a significant fire risk, necessitating effective detection and prevention measures. The early detection of fire can be detected using various ML models. The detection of fire through ML models involves data collection, preparation, model creation, and performance evaluation. Three models were analyzed: Random Forest, XGBoost, and Support Vector Classifier (SVC). The SVC demonstrated an accuracy rate of ${9 5 \%}$, while Random Forest and XGBoost achieved near-perfect accuracy. The results highlight the effectiveness of AI-driven techniques in early fire detection, showing great potential for real-world applications and improved safety protocols through automated insights. Future research will focus on refining these models for large-scale real-world deployment.