A Novel Weighted Support Vector Machine Based on Particle Swarm Optimization for Gene Selection and Tumor Classification

Table 4

Classification accuracy of our method with other methods from literature (under 10-fold cross validation).

(Authors, year)	Method	Leukemia		Colon
		Acc (%)	S. G.	Acc (%)	S. G.
(Ruiz et al., 2006) [24]	NB-FCBF	95.9	48.5	77.6	14.6
(Shen et al., 2007) [8]	PSOSVM	N. C.	N. C.	91.67	4.00
(Li et al., 2008) [15]	Single PSO	94.6	22.3	87.1	19.8
(Li et al., 2008) [15]	Single GA	94.6	23.1	87.1	17.5
(Li et al., 2008) [17]	Hybrid PSO/GA	97.2	18.7	91.90	18.00
(Shen et al., 2008) [13]	HPSOTS	98.61	7.00	93.32	8.00
(Abdi et al., 2012) [18]	mRMR-PSO-WSVM	98.74	4.1	93.55	6.8

^*S. G. and N. C. denote selected genes and not considered, respectively.