Fuzzy Weighted Least Squares Support Vector Regression with Data Reduction for Nonlinear System Modeling

Table 12

Comparison results of the selected different overlap factor for our approach are shown for Example 4, where represents the number of SRMs, L-T the computational time for constructing all SRMs, and the number of data points for each training subset.

The total training data points =296, =2.8					RMSE	RMSE	L-T
					(Training)	(Testing)	(second)
=4
	162	197	194	160	0.2236	0.4429	0.2969
	155	190	190	161	0.2801	0.2374	0.3438
	196	191	163	148	0.2883	0.2004	0.3750
	146	162	196	176	0.2563	0.3282	0.2656
	197	162	160	194	0.2236	0.4429	0.3125
					0.2544	0.3304	1.5938
The total training data points =296, =3.5					RMSE	RMSE	L-T
					(Training)	(Testing)	(second)
=4
	208	193	219	217	0.2151	0.4381	0.4219
	215	213	197	200	0.2637	0.2728	0.3125
	218	222	200	192	0.2761	0.2066	0.3906
	196	183	218	205	0.2465	0.3437	0.3438
	227	219	208	193	0.2151	0.4381	0.2969
					0.2433	0.3399	1.7747