Globality-Locality Preserving Maximum Variance Extreme Learning Machine

Table 2

Average (sum average recognition rate of different training samples per subject) recognition rate of different ELM algorithms on image datasets (%).

Dataset	ELM	MCVELM	GELM	RDELM	GLELM
	Average	Average	Average	Average	Average
ORL	92.29	92.23	93.37	91.95	93.87
Yale	67.98	67.63	71.08	67.07	75.05
Yale B	95.74	96.04	97.05	97.17	98.15
MNIST	70.00	68.34	78.36	76.90	83.49
COIL20	90.74	93.35	94.33	94.26	95.23
USPS	87.08	89.37	91.10	91.18	92.84
Average	83.97	84.49	87.55	86.42	89.77