Table 2 Summarized information for real data.

From: Ground truth clustering is not the optimum clustering

Instances

Type

n

m

\(k_{true}\)

Measuring type

Dermatology

Real

358

35

6

Mixed

Ecoli

Real

336

8

8

Numerical

Glass

Real

214

10

7

Numerical

Haberman

Real

306

4

2

Numerical

Heart-statlog

Real

270

14

2

Mixed

Iono

Real

351

35

2

Numerical

Iris

Real

150

5

3

Numerical

Sonar

Real

208

61

2

Numerical

Tae

Real

151

6

3

Mixed

Thy

Real

215

6

3

Numerical

Wine

Real

178

14

3

Numerical

Zoo

Real

101

17

7

Categorical