An Optimal Credit Scoring Model Based on the Maximum Default Identification Ability for Chinese Small Business

Table 7

The 4 types of objective weighting results.

(1) Indicator	Entropy weight method		Mean square deviation method		Wilks’ Lambda method		ROC curve
	(2) Entropy value ei	(3) Weight	(4) Mean square deviation si	(5) Weight	(6) χ^2 statistics value	(7) Weight	(8) AUC value	(9) Weight
X1 net cash flow ratio from current liabilities operating activities	0.996	0.010	0.104	0.030	12.712	0.017	0.608	0.056
X2 super-quick ratio	0.953	0.134	0.170	0.049	18.357	0.025	0.783	0.073
X3 total outstanding loans to total assets ratio	0.998	0.005	0.114	0.033	17.268	0.024	0.514	0.048
X4 net cash flow from operating activities (yuan)	0.998	0.007	0.085	0.025	25.302	0.035	0.578	0.054
X5 working capital allocation ratio	0.963	0.103	0.178	0.051	4.155	0.006	0.599	0.056
X6 retained earnings growth rate	0.993	0.019	0.119	0.034	5.041	0.007	0.566	0.052
X7 consumer price indicator	0.981	0.054	0.044	0.013	172.99	0.237	0.713	0.066
X8 controlled income of each urban resident (yuan)	0.974	0.074	0.131	0.038	107.26	0.147	0.573	0.053
X9 Engel coefficient	0.967	0.092	0.070	0.020	136.39	0.187	0.702	0.065
X10 working time in relevant industry	0.944	0.157	0.344	0.099	41.982	0.057	0.652	0.060
X11 account opening status	0.931	0.194	0.304	0.087	4.076	0.006	0.711	0.066
X12 product sales range	0.980	0.056	0.287	0.082	33.274	0.046	0.765	0.071
X13 dwelling condition	1.000	0.001	0.479	0.137	19.604	0.027	0.657	0.061
X14 working time holding the position	0.995	0.014	0.373	0.107	27.918	0.038	0.835	0.077
X15 enterprise credit in 3 years	0.999	0.002	0.351	0.101	79.264	0.108	0.823	0.076
X16 score of pledged collateral	0.972	0.080	0.330	0.095	25.060	0.034	0.708	0.066