Table 9.

Model validation with various neurons based on original dataset as input and PCA extracted components as input.

Model Characteristics	No. of Hidden Nodes	MAE	RMSE	RAE	RSE	R2
Original parameters	8	0.2060	0.2876	0.6790	0.5828	0.4371
	9	0.1904	0.2776	0.6649	0.5877	0.4372
	10	0.1885	0.2727	0.6601	0.5936	0.4463
	11	0.1897	0.2688	0.6579	0.6112	0.4737
	12	0.1769	0.2633	0.6586	0.6149	0.4850
	13	0.1734	0.2628	0.6444	0.6184	0.5165
	14	0.1725	0.2576	0.6425	0.6228	0.5371
	15	0.1704	0.2541	0.6396	0.6273	0.5526
	16	0.1692	0.2506	0.6393	0.6365	0.5584
	17	0.1691	0.2448	0.6340	0.6339	0.5806
	18	0.1647	0.2430	0.6317	0.624	0.5862
	19	0.1768	0.2568	0.6457	0.6250	0.5749
	20	0.1859	0.2618	0.6333	0.6300	0.5849
PCA extracted components	3	0.182423	0.250838	0.78808	0.701039	0.608961
	4	0.169745	0.233575	0.775804	0.696784	0.633216
	5	0.161998	0.223779	0.763809	0.687294	0.652706
	6	0.1519	0.2104	0.7619	0.6818	0.6654
	7	0.157451	0.228864	0.762694	0.695096	0.624904
	8	0.16392	0.243648	0.786798	0.702337	0.597663