A Fast Logdet Divergence Based Metric Learning Algorithm for Large Data Sets Classification

Table 3

Running time (s) comparison with the state-of-the-art metric learning methods on the normal size data sets.

Dataset	Offline LDML	Proposed	LEGO	Online ITML	POLA
Iris	0.4557	0.0306(2)	0.0287	0.0293	0.1372
Wine	0.7019	0.0434(9)	0.0387	0.0399	0.2231
Seeds	0.8957	0.0420(4)	0.0407	0.0420	0.2756
WDBC	3.4414	0.1286(4)	0.1682	0.1620	1.0273
Sonar	1.9713	0.0586(10)	0.1304	0.1291	0.6433
Ionosphere	3.3469	0.1009(13)	0.1229	0.1247	1.0725