Self-Tuning Inference Model for Settlement in Shield Tunneling: A Case Study of the Taipei Mass Rapid Transit System’s Songshan Line

Table 7

Performance evaluation summary.

Method	Average
	MAPE		RMSE		MAE		R
	Training	Testing	Training	Testing	Training	Testing	Training	Testing
PDDBS oversampling	3.000	4.820	0.785	1.537	0.530	0.841	0.996	0.986
PDDBS median sampling	2.400	3.670	0.905	1.385	0.524	0.796	0.993	0.986
SMOTE oversampling	2.610	4.150	0.846	1.444	0.447	0.665	0.995	0.988
SMOTE median sampling	2.090	3.090	0.834	1.229	0.420	0.628	0.994	0.987
Best value	2.090	3.090	0.785	1.229	0.420	0.628	0.996	0.988

The bold values indicate the best value.