The Evaluation on the Credit Risk of Enterprises with the CNN-LSTM-ATT Model

Table 3

The parameter setting of the CNN-LSTM-ATT model.

Parameter	Meaning of parameter	Value of parameter
time_series	The step length	20
behavior_num	Number of behavioral indexes	14
iB-LSTM_units	Number of neurons for LSTM	64
Kernel_size1	The size of the convolution kernel for convolution layer 1	1 × 3
Kernel_size1	The size of the convolution kernel for convolution layer 2	3 × 3
Pool	Pooling method	Max
Stride	Length of pooling or convolutional	1
Dropout	Rate of LSTM dropout	0.3
batch_size	The size of batch	128
Epoch	Number of iterations	20
Optimizer	The optimizer	SGD
Learning rate	The learning rate	0.0001