A Compound Structure for Wind Speed Forecasting Using MKLSSVM with Feature Selection and Parameter Optimization

Table 6

Forecasting results by EEMD-HGSA-LSSVM based on different kernel function.

Data set	Models	RMSE ()	MAPE ()	MAE ()
A	Model 1	0.2858	5.6439	0.2597
	Model 2	0.2957	5.8311	0.2673
	Model 3	0.2787	5.457	0.2558
B	Model 1	0.2938	6.8854	0.274
	Model 2	0.3025	7.4624	0.2786
	Model 3	0.2856	6.4077	0.2538
C	Model 1	0.2801	3.433	0.2607
	Model 2	0.2907	3.382	0.2546
	Model 3	0.2691	3.232	0.2392
D	Model 1	0.2994	4.9216	0.2819
	Model 2	0.3108	4.9586	0.2865
	Model 3	0.2903	4.5088	0.2607

Models 1, 2, and 3 denote EEMD-HGSA-LSSVM based on RBF, Poly, and multikernel function, respectively.