Applied Informatics

Table 1 The comparison of the top recognition rates and the corresponding number of features selected

From: Sparsity preserving score for feature selection

Methods	Training date
	Yale		ORL
	5	6	5	6
Data variance	0.6889 (704)	0.6800 (829)	0.9450 (2503)	0.9563 (2112)
Laplacian score	0.7111 (434)	0.7067 (952)	0.9450 (2390)	0.9563 (1901)
MCFS	0.6556 (974)	0.6933 (825)	0.9250 (1593)	0.9500 (588)
SPEC	0.7111 (836)	0.7200 (780)	0.9150 (2563)	0.9500 (2350)
SPS	0.7333 (551)	0.7333 (569)	0.9450 (2355)	0.9563 (1823)

Back to article page