Research on the Precision Impact of IV-ELM Model in the Evaluation of Debris Flow Susceptibility
Debris flows are among the most severe geological disasters,significantly impacting various construction projects and the safety of residents.Therefore,in the construction of various projects,it is crucial to avoid debris flow-prone areas and provide reliable debris flow susceptibility assessment maps for the selection of construction sites.This paper takes the debris flow disasters in Weixi County as the research subject and selects factors such as elevation,slope,engineering lithology,river density,fault density,road density,average annual rainfall(2013-2022),Normalized Difference Vegetation Index(NDVI),topography,and vegetation type as the causative factors of debris flows in the study area.The Information Value(IV)method is used to refine non-debris flow sample data to optimize the Extreme Learning Machine(ELM)model for debris flow susceptibility assessment in the study area,and the accuracy is compared with traditional ELM and IV models.The results show that the AUC(Area Under the Curve)value of the refined Extreme Learning Machine model(IV-ELM)is 0.9988,which is an improvement of 0.0893 and 0.1113 over the IV model and the ELM model,respectively.The frequency of very high susceptibility areas is increased by 0.38 and 0.29 compared to the IV model and the ELM model,respectively.
万方数据
维普
