3D Geological Modeling Method Based on Stacking and Deep Learning
We present the ET4DD (Enhanced Transformer for Drilling Data) with voting mechanism and offset-attention mechanism for three-dimensional geological modeling. To substantiate the model's efficacy, we have trained an amalgamated model that encompasses feed-forward neural networks (FNN), random forests (RF), GBDT, and XGBoost models on an identical dataset.
To run this project, you will need the following:
- Python 3.7
- NumPy (version 1.18.0 or later)
- Pandas (version 1.3.5 or later)
- Tensorflow (version 2.3.1 or later)
- PyTorch (version 1.12.0 or later)
- XGboost (version 1.6.2 or later)
- Scikit-learn (version 1.0.2 or later)