A Multikernel-Like Learning Algorithm Based on Data Probability Distribution

Table 4

Average error rates (%) of the two-class experiments on Yale-B and CMU-PIE datasets with different number of labeled samples.

Dataset	Algorithm	3 labeled samples		5 labeled samples		7 labeled samples
		Unlabeled	Test	Unlabeled	Test	Unlabeled	Test
Yale-B	MKDPD	35.2279	31.4955	30.4664	25.1786	27.3871	20.7087
	MR	36.7658	32.7455	31.7899	26.8136	28.8743	22.7087
CMU-PIE	MKDPD	23.9889	20.4167	13.6962	10.1042	8.5904	5.5754
	MR	26.0776	22.9712	15.9315	12.3611	10.0515	6.8452