[Retracted] Using an Optimized Learning Vector Quantization- (LVQ-) Based Neural Network in Accounting Fraud Recognition

Table 1

Final fraud risk identification index.

Index selection significance	Index name	Mean value		Mann–Whitney rank test		T test
		Fraud	Pairing	Z value	value	t value	value
Profitability	X1 indicator	355	186	−5.121	0.001	−3.931	0.0001
Operating capacity	X2 inventory turnover rate	29.73	965.23	−1.181	0.249	−1.731	0.0837
Solvency	X3 cash flow ratio	0.098	0.201	4.531	0.001	−2.109	0.0357
Development ability	X4 total profit	0.013	0.114	−4.274	0.001	−2.798	0.0063
Per share index	X5 before interest and tax stock income	0.340	0.442	−3.134	0.002	−2.599	0.0096
Corporate	X6 shareholding ratio of the board of supervisors	2.121	0.011	−0.242	0.002	−2.574	0.056
	X7 two-time part-time	0.645	0.125	−2.542	0.068	−1.541	0.005
	X8 management shareholding ratio	0.137	0.091	2.846	0 .004	3.026	0.002
Governance	X9 proportion of state-owned shares	0.051	0.037	−2.249	0.248	1.679	0.099
Ownership structure	X10 audit opinion type	0.054	0.056	−5.121	0.457	1.548	0.005
	X11 change of accounting firm	0.125	0.005	−2.125	0.054	1.513	0.008
Auditor relations	X12 other receivables/total assets	0.514	0.012	−2.542	0.045	1.158	0.002
Behavioral characteristics	X13 avoid ST	0.154	0.005	−1.242	0.002	1.187	0.015