4.7 Article

Three-Dimensional Mineral Prospectivity Mapping by XGBoost Modeling: A Case Study of the Lannigou Gold Deposit, China

Journal

NATURAL RESOURCES RESEARCH
Volume 31, Issue 3, Pages 1135-1156

Publisher

SPRINGER
DOI: 10.1007/s11053-022-10054-7

Keywords

3DMPM; Machine learning; XGBoost; Lannigou gold deposit

Funding

  1. National Key Research and Development Program of China [2017YFC0601502]
  2. Research on key technology of mineral prediction based on geological big data analysis [6142A01190104]

Ask authors/readers for more resources

This study applied the 3DMPM approach to explore the deep mineral resources of the Lannigou gold deposit in Guizhou. A prediction model based on the XGBoost algorithm was built, which outperformed traditional methods in terms of prediction accuracy.
Three-dimensional mineral prospectivity mapping (3DMPM) aims to explore deep mineral resources and many methods have been developed for this task in recent years. The eXtreme Gradient Boosting (XGBoost) algorithm, an improvement of the gradient boosting decision tree model, has been used widely in many fields due to its high computational efficiency and its ability to alleviate overfitting effectively. The Lannigou gold deposit in Guizhou is a well-known epithermal gold deposit in the Golden Triangle area of Guizhou, Guangxi and Yunnan, China, with potential for deep exploration. Geological data were used to establish a three-dimensional (3D) model, and subsequently a prospectivity model was built based on the metallogenic system and on geological anomaly theories. The 3D spatial reconstruction of mineralization anomalies was completed and 3D prediction layers of the ore-controlling factor were implemented to establish the basic data for the prediction model. The XGBoost classification model was proved efficient for 3DMPM, outperforming the weights of evidence method according to prediction success rate and accuracy.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available