Optimum Design of Fully Composite, Unstiffened, Built-Up, Hybrid Steel Girder Using GRG, NLR, and ANN Techniques

Table 2

Statistical analysis of collected database.

	η	α	Fcu	Ftf	Ftw	M	Hs	Asc	Asw	Ast	ts
			(t/cm^2)	(t/cm^2)	(t/cm^2)	(cm.t)	(cm)	(cm^2)	(cm^2)	(cm^2)	(cm)
Training set
Max	1.50	1.00	0.45	6.90	6.90	50000	175.3	152.4	250.4	152.7	42.0
Min	1.00	0.00	0.25	2.40	2.40	1500	20.7	0.7	5.9	3.3	10.0
Avg	1.14	0.43	0.33	4.54	3.93	16333	64.0	23.3	48.4	43.6	13.8
SD	0.15	0.37	0.09	1.61	1.45	16970	30.87	27.43	43.16	34.54	6.19
Var	0.13	0.85	0.26	0.36	0.37	1.04	0.48	1.18	0.89	0.79	0.45
Validation set
Max	1.50	1.00	0.45	6.90	6.90	50000	156.5	133.1	199.7	133.1	42.0
Min	1.50	0.00	0.25	2.40	2.40	1500	24.1	0.4	8.0	1.9	10.0
Avg	1.15	0.44	0.33	4.54	3.93	16333	64.1	23.8	48.6	43.8	13.8
SD	0.15	0.37	0.09	1.61	1.45	16970	30.83	28.13	43.12	34.79	6.20
Var	0.13	0.83	0.26	0.36	0.37	1.04	0.48	1.18	0.89	0.79	0.45