The Correlation-Base-Selection Algorithm for Diagnostic Schizophrenia Based on Blood-Based Gene Expression Signatures

Table 2

The evaluation results of data processing.

	LWL		BayesNet		SMO		KNN		NativeBayes		J48
	Before	After	Before	After	Before	After	Before	After	Before	After	Before	After
(%)	73.1	100.0	79.3	83.4	72.4	80.0	69.0	76.6	51.7	76.6	58.6	71.7
(%)	85.8	16.6	91.3	77.5	105.4	89.5	81.8	85.4	139.4	93.8	128.0	103.5
(%)	99.7	100.0	79.3	89.0	72.4	80.0	100.0	95.5	51.7	83.1	62.1	77.9
-measure	0.73	1.00	0.80	0.83	0.71	0.80	0.67	0.76	0.37	0.76	0.57	0.71
area	0.81	1.00	0.83	0.92	0.70	0.80	0.84	0.81	0.48	0.82	0.51	0.72
area	0.80	1.00	0.79	0.92	0.66	0.74	0.84	0.79	0.49	0.81	0.53	0.67

CCI, correctly classified instances; RRSE, Root Relative Squared Error; CC, Coverage of Cases (0.95 level). ROC, Receiver Operating Characteristic; PRC, Precision and Recall Curve.