Twin Support Vector Machine Method for Identification of Wiener Models

Table 6

SVR: root mean square errors between true and estimated output in Example 2 for test data set using various filter poles and filter orders. The other hyperparameters were fixed as = , = , and .

				RMSE	CPU time
0.1	1	100	1	0.4533	33.3899
			2	0.4291	27.5467
			3	0.3944	29.6294
			4	0.3655	27.5746
			5	0.3074	29.2262
			6	0.2651	32.6605
			7	0.2318	29.8089
			8	0.1926	31.2993
			9	0.1498	27.3892
			10	0.1483	26.7030
			11	0.1472	28.4650
			12	0.1466	26.9067
			13	0.1455	22.8941
			14	0.1444	22.9346
			15	0.1439	22.8492
			16	0.1433	22.1835
			17	0.1435	23.0951
0.2			16	0.1429	28.7001
0.3				0.1403	28.4638
0.4				0.1371	27.4459
0.5				0.1345	28.8956
0.6				0.1364	24.4976