Modern Machine Learning Techniques for Univariate Tunnel Settlement Forecasting: A Comparative Study

Table 1

Prediction results of all measurement points.

Point number	PSO-SVR			PSO-BPNN			PSO-ELM
	MSE	RMSE	MAPE %	MSE	RMSE	MAPE %	MSE	RMSE	MAPE %
184	0.000158	0.012562	0.653395	0.072721	0.269669	0.411474	0.246769	0.496759	1.412007
191	0.000381	0.019515	0.120525	0.02051	0.143215	0.611794	0.018798	0.137107	0.642722
192	0.000242	0.015568	1.153007	0.05744	0.239667	11.00573	0.014357	0.119823	3.166807
220	0.002564	0.050634	0.039021	0.398009	0.630879	5.88052	0.39639	0.629595	5.869113
230	0.000128	0.011333	0.141477	0.134172	0.366295	0.757707	0.117423	0.342671	0.969642
554	0.003051	0.055235	1.186884	0.254399	0.504379	1.491434	0.560271	0.748513	2.346384
569	0.001809	0.042533	1.441162	0.378479	0.615206	2.444453	1.046243	1.02286	3.023007
570	0.002652	0.051502	1.116073	0.419282	0.64752	1.932607	0.509027	0.713461	2.194194
571	0.001287	0.035874	0.351681	0.570813	0.755522	0.957628	0.794074	0.891108	1.856434
580	0.003629	0.060241	2.490363	1.560899	1.24936	11.4861	0.509974	0.714125	8.178366