Skip to main content

Table 4 Results for Clest

From: Computational cluster validation for microarray data analysis: experimental assessment of Clest, Consensus Clustering, Figure of Merit, Gap Statistics and Model Explorer

 

Precision

Timing

 

CNS Rat

Leukemia

NCI60

Lymphoma

Yeast

CNS Rat

Clest-FM-K-means-R

8

1.2 × 106

Clest-FM-K-means-A

18

7

12

15

13

1.4 × 106

Clest-FM-K-means-C

12

5

12

11

1.5 × 106

Clest-FM-K-means-S

24

8

13

15

1

1.8 × 106

Clest-FM-Hier-A

10

6

10

13

24

1.1 × 106

Clest-FM-Hier-C

10

15

8

1.1 × 106

Clest-FM-Hier-S

20

10

15

15

1

1.1 × 106

Clest-Adj-K-means-R

3

2

1.1 × 106

Clest-Adj-K-means-A

12

3

1.4 × 106

Clest-Adj-K-means-C

9

2

1.4 × 106

Clest-Adj-K-means-S

20

6

13

6

10

1.8 × 106

Clest-Adj-Hier-A

13

3

11

1.1 × 106

Clest-Adj-Hier-C

9

2

1.1 × 106

Clest-Adj-Hier-S

4

7

7

26

1.1 × 106

Clest-F-K-means-R

15

1.2 × 106

Clest-F-K-means-A

8

6

10

14

11

1.4 × 106

Clest-F-K-means-C

9

5

12

1.5 × 106

Clest-F-K-means-S

21

10

15

15

1

1.8 × 106

Clest-F-Hier-A

7

10

15

27

1.1 × 106

Clest-F-Hier-C

9

13

1.1 × 106

Clest-F-Hier-S

28

10

15

15

1

1.1 × 106

Gold solution

6

3

8

3

5

-

  1. A summary of the results for Clest on all algorithms and the first four datasets, with use of three external indexes. The columns under the label precision indicate the number of clusters predicted by Clest, while the remaining one indicates the timing in milliseconds for the execution of the corresponding experiment. For PBM, the experiments were terminated due to their high computational demand (weeks to complete). Therefore, the resulting column is omitted from the table. For the Leukemia, NCI60 and Lymphoma datasets, the timing experiments are not reported because incomparable with those of CNS Rat and of the other measures. The corresponding columns are eliminated from the table.