Software Defect Prediction through Neural Network and Feature Selections

Table 10

Comparison of RBF results with other previous methods before feature selection for KC3 data set.

Source	Algorithm	F-measure	Accuracy
[43]	J star	N/A	72.06%
[47]	Naive Bayes	0.91	86.17%
	MLP	0.94	90.04%
	SVM	0.95	90.02%
	RBF	0.90	89.78%
[31]	RF	0.71	82.80%
	DS	0.78	78.01%
	SVM	0.78	79.01%
	LR	0.79	77.01%
This research	RBF	0.85	78.25