Skip to main content

Table 4 Predicted percentage yields by ANN and RSM models along with absolute deviation, R2and AAD

From: Comparison of estimation capabilities of response surface methodology (RSM) with artificial neural network (ANN) in lipase-catalyzed synthesis of palm-based wax ester

ANN Predicted yield (%)

ANN Absolute deviation

RSM Predicted yield (%)

RSM Absolute deviation

30.19981

6.36E-06

29.4

0.027211

32.89836

4.98E-05

34.4

0.043605

75.49839

2.13E-05

75.8

0.003958

41.79979

5.05E-06

42.8

0.023364

75.19847

2.04E-05

78

0.035897

72.89809

2.62E-05

72.5

0.005517

55.59765

4.22E-05

54.1

0.027726

80.79801

2.47E-05

80.3

0.006227

78.69785

2.74E-05

79.1

0.005057

60.59789

3.48E-05

59.2

0.023649

72.39832

2.33E-05

72.3

0.001383

68.59837

2.38E-05

69.3

0.010101

48.49916

1.73E-05

46.6

0.040773

47.99756

5.09E-05

48.6

0.012346

19.99912

4.42E-05

20.2

0.009901

79.99778

2.77E-05

81.9

0.023199

70.19804

2.79E-05

68.9

0.018868

72.49749

3.46E-05

72.5

0

70.09794

2.94E-05

67.6

0.036982

81.99761

2.92E-05

83.1

0.013237

81.86748

3.08E-05

81.9

0.000366

74.07436

0.014969

78.0

0.037234

45.0392

0.010128

45.9

0.00879

46.27601

0.00165

48.0

0.038961

56.77555

0.02483

55.0

0.00722

  1. ANN training set R2 = 1
  2. ANN training set AAD (%) = 0.002844
  3. RSM R2 = 0.999619
  4. RSM AAD (%) = 0.0256
  5. ANN testing set R2 = 0.994122
  6. ANN testing set AAD (%) = 1.289405
  7. ANN training set: normal and italic (center points) numbers
  8. ANN testing set: bold numbers