Paper Title
UTILIZING ARTIFICIAL INTELLIGENCE TECHNIQUES FOR FIRE BEHAVIOR MODELING IN TASMANIAN BUTTON GRASS MOORLANDS

Abstract
Abstract - Wildfires in Tasmanian Button grass Moorlands are a major concern due to their frequency and detrimental impacts. Accurate modeling of fire spread rates in these areas is crucial for developing effective fire management strategies. This study develops an XG Boost model to predict fire spread rates using fuel moisture, wind speed, and fuel age as inputs. The model is trained on 45 data records and tested on 9 records. Results show that the benchmark statistical model outperforms XG Boost in the test set in terms of error metrics, likely due to XG Boost over fitting on the small training set. Wind speed is identified as the most influential variable affecting fire spread rate, followed by fuel moisture. More data could improve the generalization ability of XG Boost and fire spread predictions in Button grass Moorlands. Keywords - Wildfire Modeling, Fire Spread, Machine Learning, XG boost, Feature Importance, Fire Management.