Algorithm 1. FA-Based Feature Optimization.

Start Step 1:$\mathrm{Define} \mathrm{fitness} \mathrm{function}: h\left(x\right), x=\left(x_1, x_2, \dots , x_d\right)$ Step 2:$\mathrm{Generate} \mathrm{initial} \mathrm{population} \mathrm{of} \mathrm{fireflies} P_l$, where $l=1, 2, 3, \dots , n$ Step 3:$\mathrm{Compute} \mathrm{Brightness} Y\left(s\right)=Y_0{e}^{-\delta s}$ Step 4: Define Absorption Coefficient $\delta$   - $While(z<MaxGeneration)$     - $for l=1:𝕟$       - $for m=1:l$         - $if (Y_m>Y_l)$         - Vary attractiveness with distance $s$ via $e^{-\delta s}$         - Move firefly $l$ towards $m$ using $P_l^{z+1}=T_0{e}^{-\delta s_{lm}^2}\left(P_m^z-P_l^z\right)+\phi \left(Rand-0.5\right)$         - Evaluate new solutions and update brightness         - $end if$       - $end for m$     - $end for l$     - Find the latest best Firefly     - Entropy-based activation is applied $\mathrm{H}\left(\mathsf{\Phi }\right)$     - Best Optimal Features are Selected $\widetilde{{\mathsf{\Phi }}_N^k}$   - $end While$   - Processing results and visualization End