Skip to main content

Table 1 The detailed predictions on the Carcinomas dataset. Carcinomas dataset: testing result of the Disc-F-test-KNN classifier trained on the 100 samples by setting 160 gene clusters to select 80 genes. 72 out of the 74 predictions are correct (in bold).

From: Selecting dissimilar genes for multi-class classification, an application in cancer subtyping

# Samples P B C G K LI O LA LS
Prostate (P) 16 16         
Breast (B) 14   13 1       
Colorectal (C) 12    12       
Gastroesophagus (G) 1     0   1    
Kidney (K) 1      1     
Liver (LI) 1       1    
Ovary (O) 18        18   
Lung Adeno. (LA) 5         5  
Lung Squamous (LS) 6          6