|
Cluster size
|
Mean
|
SD
|
Min
|
Max
|
|---|
|
Leukemia
|
5.855
|
2.910
|
1
|
23
|
|
Breast cancer
|
4.344
|
2.062
|
1
|
13
|
|
Prostate
|
6.327
|
2.373
|
2
|
17
|
|
Colon
|
6.642
|
2.733
|
2
|
20
|
|
SRBCT
|
4.739
|
1.816
|
1
|
14
|
|
Lymphoma
|
5.485
|
2.679
|
1
|
16
|
|
Brain
|
6.094
|
2.751
|
1
|
19
|
|
NCI
|
6.174
|
2.930
|
1
|
20
|
- Variability in size of clusters that have been shaped with the supervised algorithm, based on 1000 bootstrap replicates. Leukemia stands for distinction between AML and ALL; in the breast cancer data, the separation of the ER receptor status has been analyzed; prostate and colon stand for discrimination of normal versus tumorous tissue; in the SRBCT dataset, the Ewing family of tumors was separated against three other phenotypes; for the lymphoma dataset discrimination of DLBCL against FL and CLL was considered; in the brain tumor dataset AT/RTs were discriminated from four further malignancies; and in the NCI dataset, leukemia was separated against seven other cancers. The presented figures for the four multiclass datasets are representative for all their binary distinctions between a tumor type against all others.
