Table 4.

ANOVA for predicted RSM model

Source	Sum of squares	DOF	Mean square	F value	P value
Model	64.67	23	2.81	121.20	<0.0001
x 1: Number of neurons	12.72	1	12.72	548.38	<0.0001
x 2: Learning rate	0.43	1	0.43	18.53	0.0002
x 3: Momentum constant	0.02	1	0.02	0.95	0.3397
x 4: Training epoch	1.37	1	1.37	59.08	<0.0001
x 5: Number of training run	0.02	1	0.02	0.81	0.3757
x 1 x 2	0.00	1	0.00	0.11	0.7425
x 1 x 3	0.26	1	0.26	11.04	0.0027
x 1 x 4	0.59	1	0.59	25.41	<0.0001
x 1 x 5	0.44	1	0.44	18.87	0.0002
x 2 x 3	0.09	1	0.09	3.97	0.0571
x 2 x 4	0.03	1	0.03	1.42	0.2446
x 3 x 4	0.23	1	0.23	9.71	0.0044
x 3 x 5	0.15	1	0.15	6.57	0.0165
x 4 x 5	1.01	1	1.01	43.74	<0.0001
x 21	0.36	1	0.36	15.72	0.0005
x 22	0.45	1	0.45	19.40	0.0002
x 24	1.66	1	1.66	71.73	<0.0001
x 25	0.11	1	0.11	4.68	0.0400
x 1 x 2 x 4	1.33	1	1.33	57.53	<0.0001
x 1 x 3 x 4	0.09	1	0.09	3.71	0.0651
$x_2^2$ x 2	0.15	1	0.15	6.29	0.0187
$x_2^2$ x 4	0.24	1	0.24	10.54	0.0032
$x_2^2$ x 5	0.26	1	0.26	11.03	0.0027
Residual	0.60	26	0.02
Lack of fit	0.35	19	0.02	0.52	0.8772
Pure error	0.25	7	0.04
Cor total	65.28	49