Skip to main content

Science – Society – Technology

Table 5 Ranking score and optimal selection of the eight ANN architectures using the MCDA method

From: GaS_GeoT: A computer program for an effective use of newly improved gas geothermometers in predicting reliable geothermal reservoir temperatures

Ranking/best selection Neural network model ANN architecture Score Score average
Scenario 1 Scenario 2 Scenario 3 Scenario 4
1 ANN-22a [3-34-1] 8.45 8.12 8.58 8.74 8.47
2 ANN-12a [4-13-1] 8.38 7.98 8.68 8.57 8.40
3 ANN-25a [4-10-1] 8.23 7.94 8.37 8.45 8.25
4 ANN-13a [5-9-1] 8.18 7.78 8.50 8.36 8.20
5 ANN-23a [3-15-1] 8.12 7.76 8.39 8.31 8.15
6 ANN-7a [3-18-1] 8.11 7.68 8.45 8.32 8.14
7 ANN-9a [3-15-1] 8.04 7.63 8.37 8.21 8.06
8 ANN-20a [3-18-1] 8.02 7.68 8.31 8.15 8.04
9 ANN-24 [3-14-1] 7.97 7.62 8.23 8.14 7.99
10 ANN-10 [3-9-1] 7.95 7.55 8.29 8.10 7.97
11 ANN-8 [3-9-1] 7.76 7.33 8.13 7.92 7.78
12 ANN-11 [3-9-1] 7.74 7.32 8.11 7.90 7.77
13 ANN-6 [3-12-1] 7.49 7.09 7.87 7.60 7.51
14 ANN-21 [3-6-1] 7.33 6.96 7.71 7.42 7.36
15 ANN-37 [3-9-1] 7.22 7.44 7.12 7.04 7.21
16 ANN-19 [3-27-1] 6.89 6.86 6.68 7.16 6.90
17 ANN-39 [5-6-1] 6.86 7.22 6.75 6.50 6.83
18 ANN-34 [3-8-1] 6.84 7.24 6.75 6.43 6.82
19 ANN-33 [3-8-1] 6.46 7.19 6.05 5.95 6.41
20 ANN-38 [4-8-1] 6.06 6.82 5.55 5.61 6.01
21 ANN-15 [1-34-1] 5.95 5.53 6.47 5.95 5.97
22 ANN-32 [3-8-1] 5.98 6.58 5.24 5.97 5.94
23 ANN-36 [3-7-1] 5.96 6.54 5.76 5.44 5.93
24 ANN-14 [1-33-1] 5.87 5.52 6.31 5.87 5.89
25 ANN-28 [1-14-1] 5.88 6.06 6.08 5.45 5.87
26 ANN-1 [1-32-1] 5.69 5.23 6.25 5.70 5.72
27 ANN-2 [1-3-1] 5.62 5.18 6.18 5.61 5.65
28 ANN-3 [1-13-1] 5.56 5.10 6.13 5.56 5.59
29 ANN-26 [5-5-1] 5.56 5.95 4.56 6.06 5.53
30 ANN-16 [1-16-1] 5.42 5.02 5.97 5.36 5.44
31 ANN-35 [3-11-1] 5.49 6.44 5.02 4.79 5.43
32 ANN-5 [2-28-1] 5.06 4.47 5.71 5.15 5.10
33 ANN-18 [2-34-1] 4.98 4.47 5.61 4.98 5.01
34 ANN-31 [2-9-1] 4.67 5.37 4.47 4.00 4.63
35 ANN-30 [1-8-1] 4.42 4.57 4.86 3.79 4.41
36 ANN-27 [1-10-1] 4.20 4.74 3.83 3.89 4.16
37 ANN-17 [1-25-1] 3.51 2.90 4.31 3.46 3.54
38 ANN-4 [1-26-1] 3.47 2.81 4.30 3.47 3.51
39 ANN-29 [1-4-1] 3.30 3.23 3.96 2.73 3.31
  1. aNeural network models that represent the new gas geothermometer equations proposed in this work