Default Risk Prediction of Enterprises Based on Convolutional Neural Network in the Age of Big Data: Analysis from the Viewpoint of Different Balance Ratios

Table 3

The parameters of algorithms.

Algorithm	Parameter	Value
KNN	K-neighbors	5
DT	Max depth	12
	Split criterion	Gini
SVM	Kernel function	Rbf
	Kernel scale	2
LR	Objective function minimization	SGD
	Learning rate	10^–3
	Iteration	10^4
CNN	Epoch	10–100
	Initial learning rate(AdaGrad)	2 × 10^−6
	Batch size	32–256
	Padding mode	Same
	Kernel size	1  7
SMOTE	K-neighbors	5