Using landslide-inventory mapping for a combined bagged-trees and logistic-regression approach to determining landslide susceptibility in eastern Kentucky, United States

The statistical data used in the combined machine-learning functions are available here as csv files. This includes all geomorphic data at the determined radial buffer size (36 variables), the results of the bagged trees function (12 variables), and the bagged trees resulting feature importance data.

组合机器学习函数所使用的统计数据，已以CSV文件的形式在此处提供。此类数据涵盖经确定的径向缓冲区尺寸下的全部地貌数据（共计36个变量）、装袋树（bagged trees）函数的运行结果（共计12个变量），以及装袋树模型生成的特征重要性数据。