Large-Scale Landslide Susceptibility Mapping Using an Integrated Machine Learning Model: A Case Study in the Lvliang Mountains of China

Integration of different models may improve the performance landslide susceptibility assessment, but few studies have tested it. The present study aims at exploring way to integrating and comparing results among integrated individual models. Our objective is answer this question: Will model higher accuracy compared with model? Lvliang mountains area, a landslide-prone area in China, was taken as ten factors were considered influencing system. Three basic machine learning (the back propagation (BP), support vector (SVM), random forest (RF) models) by an function where weight coefficients computed gray wolf optimization (GWO) algorithm. 80 20% data randomly selected training testing samples, respectively, maps generated based on GIS platform. illustrated that expressed under receiver operating characteristic curve (AUC) BP-SVM-RF highest (0.7898), which better than BP (0.6929), SVM (0.6582), RF (0.7258), BP-SVM (0.7360), BP-RF (0.7569), SVM-RF (0.7298). experimental authenticated effectiveness method, can be reliable for regional assessment area. Moreover, proposed procedure good option integrate seek “optimal” result.