Table 2.
Summary of facies prediction using different supervised machine learning algorithms and AutoML.
ML algorithm | Facies | Precision | Recall | F1-score |
---|---|---|---|---|
Logistic regression | Sand | 0.71 | 0.68 | 0.7 |
Shaly Sand | 0.44 | 0.44 | 0.44 | |
Shale | 0.4 | 0.29 | 0.33 | |
Coal | 0.56 | 0.73 | 0.63 | |
Macro average | 0.53 | 0.53 | 0.53 | |
Weighted average | 0.53 | 0.54 | 0.53 | |
Gradient boosting machine | Sand | 0.92 | 0.85 | 0.88 |
Shaly sand | 0.94 | 0.79 | 0.86 | |
Shale | 0.67 | 0.95 | 0.78 | |
Coal | 0.92 | 0.76 | 0.83 | |
Macro average | 0.86 | 0.84 | 0.84 | |
Weighted average | 0.86 | 0.83 | 0.84 | |
AutoML_GBM | Sand | 0.98 | 0.97 | 0.98 |
Shaly Sand | 0.95 | 0.98 | 0.97 | |
Shale | 0.99 | 0.99 | 0.99 | |
Coal | 0.99 | 0.98 | 0.99 | |
Macro average | 0.98 | 0.98 | 0.98 | |
Weighted average | 0.98 | 0.98 | 0.98 |