Algorithm 3 The HCN model for classification.

Require: $S_Q$, two-dimensional spectrum. Ensure: $ACC$, accuracy of HCN classification.  1: for $i=1,2,\cdots ,N_{epochs}$ do  2:       for $t=1,2,\cdots ,n$ do  3:             HCN ←${\left(S_Q\right)}_t^{N_{train}}$, training HCN models.  4:             Save training parameters.  5:       end for  6:       for $t=1,2,\cdots ,n$ do  7:             HCN ←${\left(S_Q\right)}_t^{N_{test}}$, input testing dataset to the trained HCN model.  8:             $ACC$← HCN, calculating the accuracy of model judgments.  9:       end for  10:     Take the average of the accuracy rate of each epoch.  11: end for