New Fuzzy Support Vector Machine for the Class Imbalance Problem in Medical Datasets Classification

Table 2

Comparison of the classification results (%) on breast dataset.

	Method	Sensitivity	Specificity	Accuracy
Linear	FSVM	95.87±0.017	95.04±0.043	95.58±0.035
	SVDD	97.71±0.065	90.90±0.013	95.28±0.052
	FSVM-CIL	95.87±0.024	95.87±0.015	95.81±0.028
	WCS-FSVM	96.33±0.067	95.04±0.056	95.87±0.047
	FSVM-CIPlin	96.98±0.039	96.49±0.022	𝟗𝟔.𝟕𝟔±𝟎.𝟎𝟒𝟎
	FSVM-CIPexp	96.68±0.011	96.69±0.042	𝟗𝟔.𝟕𝟔±𝟎.𝟎𝟑𝟕
Gaussian kernel	FSVM	96.33±0.023	95.87±0.051	96.17±0.050
	SVDD	97.30±0.065	91.25±0.013	95.44±0.052
	FSVM-CIL	96.79±0.059	95.87±0.042	96.46±0.055
	WCS-FSVM	96.97±0.030	96.69±0.093	96.76±0.067
	FSVM-CIPlin	97.25±0.055	96.29±0.032	97.05±0.042
	FSVM-CIPexp	97.25±0.055	97.52±0.045	𝟗𝟕.𝟑𝟒±𝟎.𝟎𝟑𝟑