This project aims to develop a wildfire prediction model tailored to the Kelowna region in British Columbia, Canada, spanning the years 2017-2022. Our work focuses on harnessing past weather parameters to forecast the hectares of area burned across 36 specific zones in Kelowna on a quarterly basis. Leveraging the XGBoost Regression model, we have achieved commendable accuracy in predicting wildfire outcomes for the majority of cases. However, accurately anticipating outliers, particularly those associated with significant fires in the region, remains a persistent challenge.
