Algorithm 2: CW-RNN Network Pseudocode

# CW-RNN network parameter initialization num_hidden_units = … learning_rate = … num_iterations = … batch_size = … # CW-RNN network model construction 5.model = build_cw_rnn_model(num_hidden_units, learning_rate) # CW-RNN network training 6.for iteration in range(num_iterations): # Get the current training batch data 7.batch_data = get_next_batch(batch_size) # forward pass 8.predictions = model.forward_pass(batch_data) # Compute the loss function 9.loss = calculate_loss(predictions, batch_data) # Backpropagation 10.gradients = model.backward_pass(loss) # Update network parameters 11.model.update_parameters(gradients) # CW-RNN network prediction 12.test_data = load_test_data() 13.predictions = model.predict(test_data)