A Novel Hybrid CNN-LSTM Scheme for Nitrogen Oxide Emission Prediction in FCC Unit

Algorithm 1

Pseudocode of CNN-LSTM.

Input: train_x, train_y
	Hyper-parameters: filters, kernel_size, pool_size, batch_size, rate
	Initialize ()
	Normalization (train_x, train_y)
	//The first convolution layer
	1^st ConV_model = Sequential ([Convolution2D (filters, kernel_size, name = “Conv2D_1”), MaxPooling2D (pool_size), Flatten (), Dense (units, activation), Dropout (rate), Dense (units, activation)])
	1^st ConV_model.compile (loss_function, optimizer)
	1^st ConV_model.fit (train_x, train_y, epochs, batch_size)
	//Extract the feature map
	1^st ConV_feature_model = Model (inputs, 1^st ConV_model.get_layer (“Conv2D_1”).output)
	1^st ConV_feature_output = 1^st ConV_feature_model.predict (train_x)
	//LSTM layer
	reshape (1^st ConV_feature_output)
	LSTM_model = Sequential (LSTM (units, activation, recurrent_activation), Dense (units, activation))
	LSTM_model.compile (loss_function, optimizer)
	LSTM_model.fit (1^st ConV_feature_output, train_y, epochs, batch_size)