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

Table 4

Clustering results of k-means and k-means2o on synthetic datasets.

Dataset	ACC		ARI		NMI		FMI
	k-	k-2o	k-	k-2o	k-	k-2o	k-	k-2o
Aggregation	0.7855	0.8680	0.7624	0.7438	0.8792	0.8373	0.8159	0.7992
CirclesA3	0.5833	0.5866	0.1283	0.1311	0.3616	0.3633	0.4903	0.4915
D31	0.9764	0.9290	0.9522	0.9059	0.9669	0.9498	0.9538	0.9090
Flame	0.8375	0.8375	0.4534	0.4534	0.3987	0.3987	0.7363	0.7363
Jain	0.7855	0.7855	0.3241	0.3241	0.3690	0.3690	0.7005	0.7005
K1	0.7824	0.9923	0.7318	0.9809	0.8160	0.9765	0.7962	0.9847
R15	0.9966	0.9966	0.9927	0.9927	0.9942	0.9942	0.9932	0.9932
S1	0.9938	0.9936	0.9867	0.9863	0.9866	0.9861	0.9876	0.9872
S3	0.8568	0.8568	0.7270	0.7270	0.7959	0.7962	0.7453	0.7453
Spiral	0.3461	0.3461	0.0000	0.0000	0.0007	0.0005	0.3276	0.3277
Maximum	7	8	8	7	7	6	7	7

The best clustering performance evaluation values are shown in bold.