Open AccessApplication of GBDT for division of petroleum reservoirs
Author(s) -
Yaqiong Qin,
Zhaohui Ye,
Conghui Zhang
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1437/1/012050
Subject(s) - computer science , division (mathematics) , boosting (machine learning) , decision tree , class (philosophy) , artificial intelligence , petroleum , alternating decision tree , macro , machine learning , artificial neural network , pattern recognition (psychology) , data mining , geology , decision tree learning , mathematics , incremental decision tree , paleontology , arithmetic , programming language
Traditional methods of dividing petroleum reservoirs are inefficient. The machine learning method has been applied in the two-class and three-class tasks of the reservoir, but there is no research on the classification of all classes. In this paper, the GBDT (Gradient Boosting Decision Tree) model is proposed for the reservoir classification problem in argillaceous sandstone areas, and the model is implemented by XGBoost algorithm. The classification effect of this model is better than that of multiple-hidden-layer neural networks, and the macro-average AUC is 0.89.