《Table 3 Linearity between parameters》

  提示：宽带有限、当前游客访问压缩模式

本系列图表出处文件名：随高清版一同展现

《GIS-based landslide susceptibility mapping using numerical risk factor bivariate model and its ensemble with linear multivariate regression and boosted regression tree algorithms》

1. 获取 高清版本忘记账户？点击这里登录

1. 下载图表忘记账户？点击这里登录

A linear test should be performed between the parameters in LSM，as a linear correlation between the parameters will decrease the prediction accuracy of the model（Chen et al.2017b）.In this study，two parameters（tolerance and variance inflation factor[VIF]）were used for the linear test.If the coefficient of tolerance is≤1 and the value of the VIF factor is≥10，then the existence of linearity between the parameters is confirmed（Chen et al.2017b，c）.The results of the linearity test between the parameters（Table 3）showed that，among the 20 parameters，five were co-linear，namely drainage density，profile curvature，total curvature，soil texture and stream power index.Therefore，these parameters were excluded due to their potential to decrease model accuracy，and the modelling was conducted using the remaining 15parameters.