Skip to main content

Table 1 Bias due to the order of classes.

From: A novel method for finding non-small cell lung cancer diagnosis biomarkers

Range/Class

C 1

C 2

C 3

R 1

4

6

30

R 2

6

30

4

R 3

0

4

16

  1. In this example, we consider the goodness ranking of a gene in a dataset with 100 patients and 3 classes of cancer sub-types. The expression values of the gene were discretized into 3 ranges R1, R2 and R3. There are 4 patients in class C1 with expression values in the range R1, 6 patients in class C2 with expression values in the range R1, and so on.