Modeling and Mapping of Flood Susceptibility at Que Son District, Quang Nam Province, Vietnam using CatBoost

Abstract In this research, the main objective is to model and map flood susceptibility in Que Son district, Quang Nam province, Vietnam using one of effective machine learning namely CatBoost. With purpose, a total 96 non-flood locations set 10 conditioning factors were collected construct geospatial database. Thereafter, Shap feature importance method was used validate select most important for modeling susceptibility, results showed that only 8 including aspect, slope, curvature, elevation, land cover, rainfall, distance rivers, Topographic Wetness Index (TWI) selected final modelling at study area. Validation also done various statistical indexes area under ROC curve (AUC). performance CatBoost (AUC = 0.96 training AUC 0.94 testing) good prediction Thus, it can be concluded valuable tool which assess other prone areas world. addition, generated from might helpful development better mitigation strategies