Reconstruct the Support Vectors to Improve LSSVM Sparseness for Mill Load Prediction

Table 3

Experimental results on benchmark data sets.

Data sets	Algorithms	RMSE	SV	Training time (s)	Testing time (s)
Chwirut1	NLSSVM	0.014 ± 0.001	150	0.043 ± 0.009	0.023 ± 0.004
	RCLSSVR	0.015 ± 0.002	12	0.241 ± 0.072	0.001 ± 0.000
	DRCLSSVR	0.016 ± 0.001	9	0.250 ± 0.076	0.001 ± 0.000
	SLSSVM	0.017 ± 0.002	42	0.484 ± 0.051	0.001 ± 0.000
	PLSSVM	0.017 ± 0.000	12	43.094 ± 1.377	0.001 ± 0.000
	IRRLSSVM	0.016 ± 0.002	16	0.313 ± 0.045	0.001 ± 0.000
Nelson	NLSSVM	0.016 ± 0.001	90	0.020 ± 0.006	0.005 ± 0.004
	RCLSSVR	0.016 ± 0.001	16	0.376 ± 0.014	0.001 ± 0.000
	DRCLSSVR	0.017 ± 0.002	11	0.412 ± 0.036	0.001 ± 0.000
	SLSSVM	0.019 ± 0.005	50	0.219 ± 0.043	0.003 ± 0.001
	PLSSVM	0.019 ± 0.001	130	12.971 ± 1.364	0.002 ± 0.000
	IRRLSSVM	0.018 ± 0.000	16	0.166 ± 0.006	0.001 ± 0.000
Boston Housing	NLSSVM	0.012 ± 0.001	400	0.062 ± 0.012	0.023 ± 0.006
	RCLSSVR	0.012 ± 0.001	66	2.660 ± 0.107	0.003 ± 0.001
	DRCLSSVR	0.014 ± 0.001	48	2.705 ± 0.030	0.002 ± 0.001
	SLSSVM	0.015 ± 0.001	223	1.680 ± 0.127	0.005 ± 0.001
	PLSSVM	0.014 ± 0.000	92	497.73 ± 16.52	0.004 ± 0.001
	IRRLSSVM	0.014 ± 0.001	62	3.680 ± 0.221	0.002 ± 0.001
Bodyfat	NLSSVM	0.004 ± 0.002	180	0.024 ± 0.000	0.004 ± 0.000
	RCLSSVR	0.004 ± 0.002	18	0.407 ± 0.153	0.001 ± 0.000
	DRCLSSVR	0.004 ± 0.002	14	0.415 ± 0.171	0.001 ± 0.000
	SLSSVM	0.006 ± 0.002	50	0.740 ± 0.054	0.001 ± 0.000
	PLSSVM	0.005 ± 0.001	28	45.904 ± 6.650	0.001 ± 0.000
	IRRLSSVM	0.004 ± 0.002	21	0.451 ± 0.049	0.001 ± 0.000
Pyrimt	NLSSVM	0.125 ± 0.027	60	0.016 ± 0.008	0.004 ± 0.001
	RCLSSVR	0.135 ± 0.007	9	0.064 ± 0.003	0.001 ± 0.000
	DRCLSSVR	0.136 ± 0.007	8	0.068 ± 0.001	0.001 ± 0.000
	SLSSVM	0.158 ± 0.071	16	0.343 ± 0.003	0.001 ± 0.000
	PLSSVM	0.167 ± 0.016	4	4.397 ± 0.162	0.001 ± 0.000
	IRRLSSVM	0.137 ± 0.001	8	0.487 ± 0.006	0.001 ± 0.001
Yacht_hydrodynamics	NLSSVM	0.006 ± 0.001	210	0.063 ± 0.011	0.026 ± 0.007
	RCLSSVR	0.006 ± 0.001	39	0.543 ± 0.175	0.002 ± 0.001
	DRCLSSVR	0.010 ± 0.002	37	0.600 ± 0.150	0.002 ± 0.001
	SLSSVM	0.107 ± 0.001	73	0.828 ± 0.212	0.003 ± 0.001
	PLSSVM	0.007 ± 0.001	62	80.567 ± 2.001	0.003 ± 0.001
	IRRLSSVM	0.012 ± 0.001	91	1.292 ± 0.306	0.004 ± 0.001
Airfoil_self_noise	NLSSVM	0.010 ± 0.006	1200	0.242 ± 0.013	0.004 ± 0.012
	RCLSSVR	0.011 ± 0.010	78	23.255 ± 3.245	0.003 ± 0.001
	DRCLSSVR	0.011 ± 0.010	75	23.506 ± 3.412	0.003 ± 0.001
	SLSSVM	0.012 ± 0.004	332	40.715 ± 4.671	0.004 ± 0.001
	PLSSVM	0.012 ± 0.003	94	2e + 04 ± 5e + 03	0.003 ± 0.001
	IRRLSSVM	0.012 ± 0.001	116	38.617 ± 2.313	0.004 ± 0.001
Motorcycle	NLSSVM	0.021 ± 0.004	100	0.015 ± 0.005	0.002 ± 0.000
	RCLSSVR	0.021 ± 0.004	10	0.072 ± 0.009	0.001 ± 0.000
	DRCLSSVR	0.023 0.004	7	0.080 ± 0.016	0.001 ± 0.000
	SLSSVM	0.032 ± 0.006	56	0.203 ± 0.073	0.002 ± 0.000
	PLSSVM	0.024 ± 0.001	10	16.366 ± 2.044	0.001 ± 0.000
	IRRLSSVM	0.022 ± 0.001	7	0.062 ± 0.012	0.001 ± 0.000
MPG	NLSSVM	0.011 ± 0.003	300	0.016 ± 0.007	0.005 ± 0.002
	RCLSSVR	0.012 ± 0.005	24	0.517 ± 0.011	0.001 ± 0.001
	DRCLSSVR	0.013 ± 0.004	14	0.591 ± 0.017	0.001 ± 0.000
	SLSSVM	0.022 ± 0.014	65	1.1600.270	0.001 ± 0.000
	PLSSVM	0.016 ± 0.002	17	228.07 ± 10.36	0.001 ± 0.000
	IRRLSSVM	0.012 ± 0.001	21	0.437 ± 0.007	0.001 ± 0.000