Algorithm 3

1: Initialize parameter $S$ = $\varnothing$,

2: For each candidate feature $f_i$ in $S$, calculate $SU$ between feature-feature and feature-class to eliminate redundant features using AMb.

3: For $i=1$to$n,j=i+1,f_i$first feature$,f_j$second feature

4:    calculate$S{U}_{i,c},S{U}_{i,j}$and$S{U}_{j,c}$

5:        $ifS{U}_{i,c}\ge S{U}_{j,c}$&$S{U}_{i,j}\ge S{U}_{j,c}$then

6:                   remove $f_j$ i.e.,$S⟵S-\left\{f\right\}$.

7:           else

8:    Insert $f_j$ into output selected features list $S⟵S+\left\{f\right\},$

9:        end

10: End of for

11: Return: The optimal subset of non-redundant genes S.