Forecasting Credit Risk of SMEs in Supply Chain Finance Using Bayesian Optimization and XGBoost

Table 3

Number of indicators and correct rate after step-by-step screening.

	All	First trim	Second trim
Removed	0	38	20
Remains	84	46	26
A0	–	76.34%	85.17%
A1	–	68.27%	75.30%
A	–	72.26%	82.83%

Note. “A0” is the correct rate of samples without credit risk, “A1” is the correct rate of samples with credit risk, and “A” is the correct rate of all samples.