Table 2.

Performance evaluation of the reliable negative sample extraction algorithms

Data set	Algorithm	FPR%	FNR%	Precision %	Recall %	F_measure %
Parkinson’s Disease	NB	37.25	4.57	95.43	89.78	92.52
	SPY	8.70	16.11	97.42	83.89	90.15
	Roc-SVM	6.52	15.00	98.08	85.00	91.07
Liver Disorders	NB	17.65	5.71	73.33	94.29	82.50
	SPY	36.14	0	40.00	100	57.14
	Roc-SVM	31.33	5.00	42.22	95.00	58.46
Cloud	NB	18.88	7.93	84.83	92.07	88.30
	SPY	9.52	14.92	92.77	85.08	88.76
	Roc-SVM	6.32	16.51	96.72	83.49	89.62
Ionosphere	NB	47.62	8.33	88.51	91.67	90.06
	SPY	26.32	6.98	94.12	93.02	93.57
	Roc-SVM	33.33	8.89	94.25	91.11	92.66
MAGIC Gamma Telescope	NB	10.49	44.44	68.18	55.56	61.22
	SPY	17.88	36.22	53.88	63.78	58.42
	Roc-SVM	6.68	47.18	77.65	52.82	62.87
Mammographic Mass	NB	7.25	33.72	85.07	66.28	74.51
	SPY	11.96	10.00	62.07	90.00	73.47
	Roc-SVM	1.95	28.57	94.34	71.43	81.30
Breast Cancer Wisconsin	NB	13.85	12.26	91.18	87.74	89.42
	SPY	9.09	10.48	94.00	89.52	91.71
	Roc-SVM	22.50	22.14	91.89	77.86	84.30
Connectionist Bench (Sonar, Mines vs. Rocks)	NB	13.85	12.26	91.18	87.74	89.42
	SPY	16.67	7.69	80.00	92.31	85.71
	Roc-SVM	22.50	22.14	91.89	77.86	84.30