Algorithm 1 Procedure of the SA-BLDA Algorithm

Require: Minimum and maximum stimulation rounds $L_{min}=4$, $L_{max}=8$; decision threshold ${\theta }_0=0.2$ Ensure: Predicted target $u_{max}$   1:Initialize $i=1$, $L_i=1$, and $\mathsf{\Delta }\theta =0$   2:Obtain the initial feature vector $f_i$ and feature matrix ${\mathbf{F}}_i$   3:while $L_i\le L_{max}$ do   4:    Update the averaged feature vector ${\mathbf{f}}_i={\displaystyle \frac{1}{i}}{\sum }_{m=1}^i{\mathbf{f}}_{i,m}$ and feature matrix ${\mathbf{F}}_i$   5:    if $L_i\ge L_{min}$ then   6:        Apply BLDA to ${\mathbf{F}}_i$ and obtain the regression score vector u   7:        Identify the maximum score $u_{max}$ and the second-largest score $u_{\sec }$   8:        Compute the confidence difference $\mathsf{\Delta }\theta =u_{max}-u_{\sec }$   9:        if $\mathsf{\Delta }\theta \ge {\theta }_0$ or $L_i=L_{max}$ then 10:           Output $u_{max}$ as the predicted target 11:           break 12:        end if 13:    end if 14:    Update the stimulation round $L_{i+1}=L_i+1$ 15:    Set $i=i+1$ 16:end while