A Fast Density Peak Clustering Method with Autoselect Cluster Centers

Table 6

Comparison of the evaluation metrics of the four algorithms on the real-world datasets.

Dataset	Algorithm	ARI	NMI	Homogeneity
Zoo	K-means	0.42069	0.64396	0.63305
	FCM	0.36469	0.62433	0.66487
	DPC	0.3056	0.42223	0.32649
	GDPC (0.718)	0.80158	0.84129	0.84208
Thyroid	K-means	0.54707	0.39504	0.40321
	FCM	0.4413	0.34344	0.37
	DPC	0.52063	0.47818	0.40368
	GDPC (0.045)	0.68386	0.55497	0.57709
Ecoli	K-means	0.43884	0.61493	0.70525
	FCM	0.34935	054486	0.6399
	DPC	0.45087	0.60743	0.63245
	GDPC (0.093)	0.63227	0.60598	0.53881
Machine	K-means	0.06195	0.28402	0.32626
	FCM	0.40954	0.44839	0.54375
	DPC	0.38929	0.43189	0.38985
	GDPC (0.106)	0.50151	0.53098	0.42842
Hayes-Roth	K-means	0.03731	0.05575	0.05575
	FCM	-0.01491	0.0	0.0
	DPC	0.00591	0.13145	0.08757
	GDPC (0.166)	0.03223	0.25694	0.50898
Sobar-72	K-means	0.26671	0.2106	0.22378
	FCM	0.26671	0.2106	0.22378
	DPC	-0.07399	0.06523	0.05014
	GDPC (0.629)	0.18569	0.31803	0.50194
Segment	K-means	0.36336	0.46123	0.44182
	FCM	0.50632	0.61017	0.60989
	DPC	0.24025	0.52623	0.40106
	GDPC (0.089)	0.43038	0.58043	0.75747
Pendigits	K-means	0.35528	0.53178	0.52341
	FCM	0.43647	0.57868	0.53033
	DPC	0.55422	0.72433	0.69383
	GDPC (0.15)	0.50352	0.60201	0.69527

The meaning of the bold value is to emphasize that the value is the best result of the four algorithms on the same dataset and metric.