Improvement of K-Means Algorithm and Its Application in Air Passenger Grouping

Table 8

Results of different initialization methods of k-means algorithm.

Dataset	ARI			NMI
		-Sampling			-Sampling
Breast-cancer	0.4914	0.4914	0.4914	0.4914	0.4914	0.4914
Banknote	0.0485	0.0485	0.0485	0.0485	0.0485	0.0485
Bupa	0.0000	0.0000	0.0000	0.0000	0.0001	0.0001
Compound	0.5328	0.5378	0.4133	0.7220	0.7191	0.6240
Ct	0.4160	0.4160	0.0000	0.3296	0.3296	0.0219
Hayes-Roth	0.0160	0.0202	0.0202	0.0250	0.0287	0.0287
Iris	0.7302	0.7302	0.7302	0.7581	0.7581	0.7581
Libras	0.3062	0.3199	0.2868	0.5896	0.6066	0.5767
Parkinsons	0.0853	0.0000	0.0001	0.0505	0.0001	0.0001
Penbased	0.5442	0.5992	0.4265	0.6835	0.6927	0.6486
Vowel	0.2180	0.2028	0.2058	0.4332	0.4141	0.4067
Waveform21	0.2536	0.2536	0.2535	0.3622	0.3622	0.3622
Waveform40	0.2516	0.2516	0.2515	0.3605	0.3605	0.3605
Wdbc	0.4914	0.4914	0.4914	0.4647	0.4647	0.4647
Wine	0.3711	0.3711	0.3711	0.4287	0.4287	0.4287
Maximum	11	13	7	11	12	9