Estimation of Soil Cohesion Using Machine Learning Method: A Random Forest Approach

Table 4

Statistical analysis of prediction results over 200 RF simulations in terms of RMSE, MAE, and R.

Datasets	Training set			Testing set
Criteria	RMSE	MAE	R	RMSE	MAE	R
Min	2.7814	2.0009	0.8702	2.6817	2.1097	0.3084
Average	3.2450	2.3702	0.8982	4.7317	3.5404	0.7140
Median	3.2635	2.3770	0.8974	4.7337	3.5521	0.7264
Max	3.5852	2.6802	0.9333	6.3457	4.8180	0.8702
SD	0.1599	0.1314	0.0103	0.6498	0.4807	0.0844
SK	−0.3224	−0.0607	0.3887	−0.0614	−0.0437	−1.3655

SD = standard deviation; SK = skewness.