Algorithm 2 DL model training

Input:  $W_0,\eta ,B,\zeta ,\rho ,\tau$

Output:  $W_n^r$

1: Divide $D_n$, into equal size B batches with feature vector x;

2: Set $W_n^r$ with initial values;

3: For each Batch;

$c_1\leftarrow$ Forward x to $Con{v}_1$;

$c_2\leftarrow$ Forward $c_1$ to $Con{v}_2$;

$\lambda \leftarrow$ Flatten $\left(c_2\right)$;

$H^{\prime }\leftarrow$ Forward $\lambda$ to $LST{M}_1$;

$\mu \leftarrow$ Forward $H^{\prime }$ to $LST{M}_2$;

$M\leftarrow$ Forward $\mu$ to $Dense$;

$\gamma \leftarrow$ Dropout $\left(M\right)$;

$\nu \leftarrow$ Forward $\gamma$ to Output(Sigmoid);

4: Compute loss function using:

$\zeta =-\frac{1}{B}{\sum }_{i=0}^1{x}_i.log{\widehat{x}}_i+(1-x_i).log(1-{\widehat{x}}_i)$;

5: Update $W_n^r$;

6: Repeat until $\zeta$ converges;