Table 2.

Statistics of neural network performance for the dataset.

Training algorithm	Size of hidden layer	Size of neuron in the hidden layer	Coefficient of correlation (R-value)								Mean squared error (MSE)
			Training data				All data
			Mean	std	Max	Min	Mean	std	Max	Min	Mean	std	Max	Min
LM	1	20	0.94055	0.001279	0.94246	0.93819	0.939135	0.000748	0.94047	0.93784	0.118962	0.004079	0.12507	0.11161
		15	0.939863	0.001341	0.94201	0.93713	0.938104	0.000677	0.93904	0.9369	0.122699	0.007997	0.13602	0.10816
		10	0.937476	0.001164	0.93964	0.93545	0.936163	0.000934	0.93758	0.93413	0.127499	0.003904	0.13295	0.11952
	2	20 by 20	0.94319	0.00064	0.94451	0.94238	0.94084	0.000596	0.94186	0.93982	0.120795	0.005911	0.13084	0.11182
		15 by 15	0.942289	0.001785	0.94571	0.93993	0.940412	0.000978	0.94267	0.93884	0.122108	0.004781	0.1301	0.11279
		10 by 10	0.941227	0.000832	0.94295	0.93967	0.939821	0.000606	0.94099	0.93897	0.118838	0.002941	0.12354	0.11267
SCG	1	20	0.917034	0.007736	0.93144	0.9006	0.916749	0.007709	0.93091	0.90041	0.157926	0.01696	0.19535	0.12816
		15	0.912231	0.006038	0.91826	0.89643	0.912168	0.005652	0.91843	0.89734	0.164146	0.010635	0.18692	0.14858
		10	0.908912	0.00523	0.9206	0.9027	0.908901	0.004398	0.91851	0.90297	0.169053	0.01151	0.19026	0.15578
	2	20 by 20	0.919298	0.004164	0.92506	0.91062	0.918904	0.003736	0.92508	0.91201	0.15345	0.006737	0.1685	0.14288
		15 by 15	0.912044	0.013226	0.92305	0.87585	0.911392	0.012733	0.92134	0.87728	0.165776	0.02495	0.23228	0.14435
		10 by 10	0.910744	0.007895	0.92179	0.89556	0.909868	0.008372	0.92073	0.89374	0.168445	0.013742	0.19757	0.1536