Abstract
Exploring more effective landslide susceptibility assessment methods play an important role in mitigating landslide effects. This paper aims to compare the performance of different popular ensemble learning models that combined with GIS system to assess coseismic landslide susceptibility in the 2017.8.8 Jiuzhaigou earthquake area. Eight influencing factors (slope, elevation, aspect, relief altitude, lithology, peak ground acceleration, distance to river, distance to fault) were considered to construct a spatial database after the Pearson correlation analysis. The 4834 landslides from inventory data are randomly divided into 70% train samples, and 30% validate samples. We construct the random forest (RF), gradient-boosting decision tree (GBDT), and adaptive boosting (AdaBoost), which all three models use decision tree model as basic unit, and utilize the receiver operating characteristic (ROC) curve, area under curve (AUC) values, Kappa value to validate the performance of three ensemble models. The results indicate that the AdaBoost model achieved the best performance (AUC= 94.4%, Kappa=0.766), outperforming the GBDT (AUC=92.5%, Kappa=0.720), RF (AUC= 93.6%, Kappa=0.756) focused on the validation data. This study can provide an insight into evaluating coseismic landslide susceptibility with high accuracy.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IOP Conference Series: Earth and Environmental Science
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
