Algorithm 1

Input: the set of all features $Y=\left\{y_1,y_2,\dots ,Y_n\right\}$ Output: a subset of features $X=\left\{x_j\right|j=1,2,3,\dots ,k;x_j\in Y\}$ Where, $k=\left(0,1,2,\dots ,n\right)$ Steps: 1. $Y_0=\{\varnothing \}$ 2. Select the best Feature X+ Update: $Y_{K+1}=Y_K+X^{+};= +1$ 3. Select the best Feature X− 4. If $J\left(Y_K-X^{-}\right)>J\left(Y_K\right)$    [$\left(X\right)$ = Criterion func.] Then, $Y_{K+1}=Y_K-X^{-};K=K+1$ go to step 3.