Algorithm 1 Algorithm for filtering based CSNN adaptive prediction.

1:Initialize the neural networks. 2:Calculate the local sample mean using Equation (1). 3:for t = 1:N do %% N is the number of batch, and the batch size equals to the number of columns in the spectral band. 4:    Select the spatial and spectral contexts $C_t$ and $C_l$ for each pixels in batch, and prepare the data pair $\left\{\left(C_t,C_l\right),s_{x,y,z}\right\}$. 5:    Extract spatial and spectral features $F_t$ and $F_l$ from the contexts using one-layer shallow neural networks. 6:    Concatenate the features: $F=\left[F_t,F_l\right]$. 7:    Predict the pixel values based on F using Equation (11). 8:    Calculate and record the prediction error $e_{x,y,z}$ for further mapping and coding. 9:    Calculate the weight updates $\Delta w$ using Equation (13). 10:    Adjusting the parameters in every batch: $w=w+\Delta w$. 11:end for