Table 2.

The results of parameters tuning in MLP, RBFNN and SVM structure.

Activation function	Training function	Structure	Test set	Training Data
			Accuracy (%)	Accuracy (%)
Tanh-Tanh	Scaled conjugate gradient (SCG)	15–27-1 MLP	93.3	100
Spread	Neurons	Structure	Test set	Training Data
495	35	RBFNN	90.6	94.8
C	γ	Structure	Test set	Training Data
200	24	SVM	93.4	97.9