Hybrid Disease Diagnosis Using Multiobjective Optimization with Evolutionary Parameter Optimization

Table 20

Performance of hybrid systems on WBC dataset.

Breast cancer
	Resampling	Without resampling	Resampling			Without resampling
			PSO	GSA	FA	PSO	GSA	FA
	Basic SVM		Parameter optimized SVM
PAC	96.14	96.85	97.00	97.00	97.00	97.00	96.85	96.85
Sensitivity	92.59	94.69	92.43	92.43	92.43	95.08	94.69	94.69
Specificity	98.03	98.02	99.55	99.55	99.55	98.02	98.02	98.02
F-measure	96.15	96.86	97.02	97.02	97.02	97.00	96.86	96.86
Recall	96.14	96.85	97.00	97.00	97.00	97.00	96.85	96.85
Precision	96.21	96.87	97.17	97.17	97.17	97.01	96.87	96.87
	Basic MLP		Parameter optimized MLP
PAC	97.14	95.28	98.00	97.71	98.00	96.57	97.00	96.42
Sensitivity	95.34	92.62	96.61	96.58	96.61	93.93	94.35	93.55
Specificity	98.06	96.70	98.70	98.28	98.70	98.01	98.45	98.00
F-measure	97.14	95.29	98.00	97.71	98.00	96.58	97.01	96.44
Recall	97.14	95.28	98.00	97.71	98.00	96.57	97.00	96.42
Precision	97.15	95.30	98.00	97.71	98.00	96.60	97.04	96.47