Differential Diagnosis Model of Hypocellular Myelodysplastic Syndrome and Aplastic Anemia Based on the Medical Big Data Platform

Table 6

The comparison of the training set evaluation index.

Aspects	Index	Logistic	Decision tree	BP neural network	SVM
Authenticity	Sensitivity (%)	68.75	98.96	84.38	69.79
	Specificity (%)	78.76	92.04	78.76	82.30
	Youden index	0.48	0.91	0.63	0.52
	+LR	3.24	12.42	3.97	3.94
	−LR	0.40	0.01	0.20	0.37
	AUC (95% CI)	0.74 (0.67, 0.80)	0.96 (0.92, 0.98)	0.82 (0.76, 0.87)	0.76 (0.70, 0.82)
Reliability	Accuracy (%)	74.16	95.22	81.34	76.56
	Kappa	0.48	0.90	0.63	0.53
Benefit	+PV (%)	73.33	91.35	77.14	77.01
	−PV (%)	74.79	99.05	85.58	76.23