Comparative Analyses of Response Surface Methodology and Artificial Neural Network on Medium Optimization for Tetraselmis sp. FTC209 Grown under Mixotrophic Condition
Table 2
Central composite design (CCD) used in RSM and ANN studies showing the actual levels and observed responses during the mixotrophic cultivation Tetraselmis sp. FTC 209.
Run
(g/L)
Lipid yield, / (mg/g dcw)
Lipid productivity, Prlipid (mg/L per day)
1
20.00
0.50
3.00
4.955
76.81
31.72
2
30.00
0.50
3.00
8.15
87.36
59.33
3
20.00
1.80
3.00
9.325
75.60
58.75
4
30.00
1.80
3.00
9.65
77.99
62.72
5
20.00
0.50
7.00
7.365
98.30
60.33
6
30.00
0.50
7.00
10.1
72.67
61.17
7
20.00
1.80
7.00
7.245
105.31
63.58
8
30.00
1.80
7.00
12.69
58.87
62.26
9
16.59
1.15
5.00
6.835
130.03
74.06
10
33.50
1.15
5.00
13.05
70.25
76.40
11
25.00
0.06
5.00
8.6
85.28
61.12
12
25.00
2.25
5.00
10.195
61.11
56.00
13
25.00
1.15
1.65
4.315
108.95
39.18
14
25.00
1.15
8.35
6.275
81.17
42.44
15
25.00
1.15
5.00
11.555
168.83
162.57
16
25.00
1.15
5.00
10.175
204.29
173.22
17
25.00
1.15
5.00
12
177.18
177.18
18
25.00
1.15
5.00
11.07
171.40
158.11
19
25.00
1.15
5.00
11.55
183.30
176.42
ANN training set: normal numbers; ANN testing set: bold numbers.