Algorithm 2 Training Process

Input:  $y$: bandwidth slice, p: Epochs, B: Batch size, X: training, g: testing, $\alpha$: learning rate, $\widehat{ \vartheta }:$ Initial Model  Output:  $\vartheta$: LSTM Model, E: Forecast Error, P: parameters  Process:  01: begin  02: for I $\leftarrow to p$  03: $\vartheta \leftarrow$forward propagate ($\widehat{ \vartheta },$ X, y, $\alpha ,B$)  04: E$\leftarrow$ Backward Propagation ($\vartheta$,g)  05: P$\leftarrow Update\left[\vartheta ,E,\alpha \right]$  06: End for