Table 6.

Grid search space for the tuning of the CNN and NN hyperparameters.

CNN model hyperparameters		Search space
Activation function		[tanh, relu]
Number of neurons on FC layer		[128, 250, 512, 1024]
Initialization mode		[uniform, glorot_normal]
Batch size		[16, 32, 64, 128]
Dropout rate		[0.01, 0.1, 0.2, 0.3, 0.4, 0.5]
Optimizer		[SGD, Adam, Nadam]
4 × L embedding	Number of filters	[16, 32, 64]
	Filter length	[1, 2, 3, 4, 5, 6, 7, 8, 9]
	Filter width	[4]
64 × L embedding	Number of filters	[16, 32, 64]
	Filter length	[3, 5, 10, 21, 31, 41, 50, 61]
	Filter width	[64]

NN model hyperparameters	Search space
Size of nodes	[1,2,3,4,5]
Activation function	[tanh, relu]
Solver	[‘lbfgs’, ‘sgd’, ‘adam’]