A New Feature Ensemble with a Multistage Classification Scheme for Breast Cancer Diagnosis

Table 12

The evaluation metrics of each classifier evaluated by elevenfold cross-validation for the three-stage study.

Classifier	SNS	SPC	PPV	NPV	FPR	FNR	FDR	FOR	ACC
LLC	87.90	93.25	90.73	94.05	6.75	12.10	9.27	5.95	92.39
FLDA	87.91	92.90	90.61	93.96	7.10	12.09	9.39	6.04	93.29
LDC	87.30	91.71	88.29	92.17	8.29	12.70	11.71	7.83	92.83
SVM (‘p’, 1)	74.85	85.29	79.66	88.45	14.71	25.15	15.43	11.55	91.80
Decision tree	85.68	91.60	87.03	92.18	8.40	14.32	12.97	7.82	87.36
Random forest	79.54	88.69	78.44	89.24	11.31	20.46	21.56	10.76	74.81
Naïve Bayes	65.23	82.04	68.49	85.73	17.96	34.77	31.51	14.27	76.00
kNN ()	63.02	77.98	54.81	78.50	22.02	36.98	45.19	21.50	50.33