Geoscience Knowledge-Guided Machine Learning for Cross-Well Lithology Recognition
The results show that combining geological knowledge with few-shot learning can improve the robustness of cross-well lithology recognition and provide a practical approach for lithology prediction in heterogeneous volcanic reservoirs.