Table 3.

The mean predicted error according to the number of hidden layer nodes by neural network

	Number of hidden layer nodes						Learning rate
	20	18	16	14	12	10
Mean square error	27.04	27.12	27.52	27.56	27.74	28.51	0.05
	26.84	26.94	27.56	27.44	27.80	28.46	0.01
	26.8	27.23	27.39	27.54	27.87	28.38	0.20
	26.94	27.37	27.38	27.51	27.86	28.37	0.50