Explainable deep learning for automatic rock classification

Dongyu Zheng, Hanting Zhong, Gustau Camps-Valls, Zhisong Cao, Xiaogang Ma, Benjamin Mills, Xiumian Hu, Mingcai Hou, Chao Ma

Abstract: Highlights•Proposed dual network achieves high accuracy (0.99) and interpretable feature extractions for sedimentary rock classification.•Regular DL models achieved high accuracy (>0.94), but relied on irrelevant features for classification.•This study emphasizes importance of interpretability and geological knowledge in developing DL models for geosciences.