Table 8 Feature importance of the input variables for models ANN-B through ANN-K
Mean(|SHAP value|) (average impact on model output magnitude) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Models | S | T | C | M | TH | b | B | TR | SN | PRS |
ANN-B | 6.1 | 6.1 | 10.1 | 8.6 | 13 | 11.9 | 14.4 | 14.8 | 15.8 | 174.6 |
ANN-C | 4.9 | 7.7 | 5.6 | 5.6 | 7.7 | 3.5 | 5.6 | 11.8 | 11.9 | 169.1 |
ANN-D | 24.1 | 18.6 | 33 | 16.3 | 25.1 | 10 | 15.7 | 33.7 | 66.4 | 75.8 |
ANN-E | 2.4 | 1.9 | 1.9 | 2.9 | 1.9 | 1.9 | 3.4 | 3.9 | 4.4 | 115.4 |
ANN-F | 2.6 | 1.3 | 4.3 | 5.4 | 5.4 | 4.3 | 3.7 | 5.7 | 22.0 | 104.1 |
ANN-G | 3.3 | 3.7 | 5.6 | 2.9 | 5.2 | 3.7 | 3.7 | 7.1 | 11.4 | 100.2 |
ANN-H | 2.3 | 3.3 | 5.0 | 4.6 | 6.9 | 6.9 | 8.4 | 11.5 | 18.1 | 70.4 |
ANN-I | 8.0 | 9.8 | 7.3 | 6.2 | 10.3 | 4.3 | 6.8 | 10.3 | 12.3 | 110.5 |
ANN-J | 8.9 | 5.8 | 6.0 | 6.0 | 6.8 | 4.5 | 8.9 | 11.1 | 14.7 | 69.1 |
ANN-K | 5.8 | 1.7 | 2.3 | 1.7 | 2.4 | 2.0 | 2.4 | 5.8 | 6.8 | 69.4 |