An Extreme Learning Machine Based on the Mixed Kernel Function of Triangular Kernel and Generalized Hermite Dirichlet Kernel

Table 7

Performance comparison of deferent algorithms in multiclass datasets.

Datasets	Algorithms
	Gauss KELM		Poly KELM		Gau-H KELM		Tri-H, KELM		Tri-H SVM
	Testing rate (%)	Training time (s)	Testing rate (%)	Training time (s)	Testing rate (%)	Training time (s)	Testing rate (%)	Training time (s)	Testing rate (%)	Training time (s)
Iris ()	98 ()	0.0005	97 ()	0.0012	98 ()	0.0038	98 ()	0.0007	96 ()	0.073
Glass ()	77.78 ()	0.0026	72.22 ()	0.004	73.61 ()	0.0036	76.37 ()	0.0031	74.58 ()	4.54
Wine ()	80 ()	0.0012	96.67 ()	0.0046	97 ()	0.0048	97.48 ()	0.0025	96.44 ()	1.41
Vehicle ()	80.5 ()	0.0163	81.21 ()	0.0076	82.62 ()	0.0206	82.62 ()	0.064	81.58 ()	1.82
Segment ()	87.4 ()	0.2002	88.44 ()	0.0834	87.27 ()	0.4636	88.68 ()	0.5265	87.51 ()	14.65
Waveform ()	86.75 ()	0.0361	84.25 ()	0.0254	86.25 ()	0.0438	87.04 ()	0.1266	85.33 ()	2.87