Training Classifiers under Covariate Shift by Constructing the Maximum Consistent Distribution Subset

Table 4

The running time and parameter values in Experiment  2.

Data set	SVM	IDC2	KLIEP
	Australian	225.6, 100, 64	209.1, 60, 64	218.9, 200, 128
Balance	262.1, 20, 128	225.2, 500, 64	278.2, 80, 16
Breast	312.1, 10, 128	288.2, 60, 128	315.5, 200, 16
Cleveland	196.2, 500, 256	160.1, 200, 1024	171.5, 80, 32
Credit	1029.2, 100, 64	899.3, 30, 256	1050.6, 20, 64
Diabetes	469.2, 200, 64	420.5, 300, 1024	425.5, 50, 16
Heart	28.8, 50, 512	22.1, 10, 16	30.3, 20, 64
Ionosphere	46.5, 10, 16	40.9, 200, 64	42.5, 80, 256
Iris	9.2, 100, 128	6.3, 60, 16	9.9, 200, 32
Liver	18.2, 60, 64	13.2, 100, 32	16.2, 60, 256
Page	32.1, 80, 1024	29.6, 1000, 16	32.6, 20, 16
Sonar	12.6, 20, 64	10.1, 200, 64	29.6, 60, 256
Thyroid	29.8, 100, 64	25.5, 400, 16	30.6, 100, 64
Vehicle	1286.2, 200, 16	1120.3, 100, 128	1225.6, 200, 16
Voting	44.6, 20, 16	38.2, 80, 512	50.5, 200, 1024
Waveform21	556.3, 600, 64	513.2, 300, 16	532.8, 100, 256
Waveform40	443.3, 800, 128	399.6, 200, 64	422.1, 600, 256
Wine	49.6, 20, 256	40.3, 100, 64	44.5, 50, 512
wdbc	266.6, 600, 64	236.2, 200, 256	299.5, 10, 128
wpbc	198.7, 10, 64	155.3, 500, 128	167.8, 200, 64