Paper Title
Analyzing the Factors Contributing to Forest Fire Risk and Utilizing Decision Tree for Prediction

Abstract
Forest Fire pose significant socio-economic and environmental consequences particularly in vulnerable areas that are prone to fire such as the Mediterranean. Their increasing intensity and frequency driven by climate change, make early and accurate predictions essential for disaster management strategies for hazard risk mitigation. This study aims to predict forest fire risk by analyzing the key factors influencing forest fire occurrence in Lebanon and evaluates their contribution to fire risk. To enhance prediction accuracy and fire management strategies, a prediction model was developed using Decision tree within the machine learning (ML) framework. Allowing for the classification and prioritization of risk factors based on their impact. Utilizing data from historical fire records, Geographic Information System (GIS), ML and remote sensing to combine various factors for developing a forest fire risk map (FFRM). The findings demonstrate the significant potential of integrating ML algorithms with real-time satellite data in capturing complex relationships between factors and fire occurrence, by providing valuable insights for early warning systems and implementing data-driven contingency plans for stakeholders. The Model achieved high prediction accuracy (0.86) and Area Under Curve (AUC = 0.896), demonstrating its potential for enhancing early warning systems and fire management strategies. Keywords - Forest Fire Occurrence, Machine Learning, GIS, Remote Sensing