Comparative Analyses of Response Surface Methodology and Artificial Neural Network on Medium Optimization for Tetraselmis sp. FTC209 Grown under Mixotrophic Condition

Table 3

ANOVA tables showing the regression model on ^abiomass concentration, ^blipid yield, and ^clipid productivity.

Source	Sum of squares	DF	Mean squares	𝐹 value	Prob > 𝐹
Biomass concentration^a
Model	115.55	9	12.84	14.20	0.0003	Significant
A (glucose)	37.52	1	37.52	41.49	0.0001
B (yeast extract)	12.12	1	12.12	13.41	0.0052
C (NaNO3)	6.09	1	6.09	6.73	0.0290
A^2	1.91	1	1.91	2.11	0.1804
B^2	2.45	1	2.45	2.70	0.1345
C^2	55.43	1	55.43	61.30	<0.0001
AB	0.05	1	0.05	0.06	0.8117
AC	2.16	1	2.16	2.39	0.1563
BC	1.90	1	1.90	2.10	0.1810
Residual	8.14	9	0.90
Lack of Fit	6.21	5	1.24	2.57	0.1907	Not significant
Pure Error	1.93	4	0.48
Cor Total	123.69	18
Lipid yield^b
Model	38640.62	9	4293.40	22.14	<0.0001	Significant
A (glucose)	1454.22	1	1454.22	7.50	0.0229
B (yeast extract)	168.75	1	168.75	0.87	0.3752
C (NaNO3)	27.36	1	27.36	0.14	0.7159
A^2	12340.99	1	12340.99	63.65	<0.0001
B^2	20415.60	1	20415.60	105.30	<0.0001
C^2	13070.60	1	13070.60	67.41	<0.0001
AB	189.90	1	189.90	0.98	0.3482
AC	1128.46	1	1128.46	5.82	0.0391
BC	4.83	1	4.83	0.02	0.8781
Residual	1744.99	9	193.89
Lack of Fit	942.45	5	188.49	0.94	0.5391	Not significant
Pure Error	802.53	4	200.63
Cor Total	40385.61	18
Lipid productivity^c
Model	48411.82	9	5379.09	64.82	<0.0001	Significant
A (glucose)	160.43	1	160.43	1.93	0.1978
B (yeast extract)	80.03	1	80.03	0.96	0.3518
C (NaNO3)	163.34	1	163.34	1.97	0.1942
A^2	14549.79	1	14549.79	175.32	<0.0001
B^2	19719.16	1	19719.16	237.61	<0.0001
C^2	26767.04	1	26767.04	322.53	<0.0001
AB	133.84	1	133.84	1.61	0.2360
AC	190.05	1	190.05	2.29	0.1645
BC	136.06	1	136.06	1.64	0.2324
Residual	746.92	9	82.99
Lack of Fit	448.49	5	89.70	1.20	0.4416	Not significant
Pure Error	298.43	4	74.61
Cor Total	49158.74	18