Investigation of Applying Machine Learning and Hyperparameter Tuned Deep Learning Approaches for Arrhythmia Detection in ECG Images

Table 4

CNN-Hyper-tuned model summary.

Layer (type)	Output shape	Number of parameters
Conv2d_4 (Conv2D)	(None, 62, 62, 16)	448
Conv2d_5 (Conv2D)	(None, 60, 60, 16)	2320
Max_pooling2d_2 (MaxPooling 2D)	(None, 15, 15, 16)	0
Dropout (dropout)	(None, 15, 15, 16)	0
Conv2d_6 (Conv2D)	(None, 13, 13, 32)	4640
Conv2d_7 (Conv2D)	(None, 11, 11, 64)	18496
Max_pooling2d_3 (MaxPooling 2D)	(None, 5, 5, 64)	0
Dropout (dropout)	(None, 5, 5, 64)	0
Flatten_1 (flatten)	(None, 1600)	0
Dense_3 (dense)	(None, 128)	204928
Dropout_2 (dropout)	(None, 128)	0
Dense_3 (dense)	(None, 6)	774
Total params: 231,606	Trainable params: 231,606	Nontrainable params: 0