A mineral deposit type classifier based on trace elements of sphalerite
Introduction
The sphalerite classifier is a classifier trained on multiple machine learning algorithms to predict the type of mineral deposit based on the trace element composition of sphalerite. This program facilitates deposit genesis identification on computers lacking a machine learning environment.
This program uses four machine learning algorithms: random forest, XGBoost, support vector machine (rbf kernel), decision tree, and linear PCA dimensionality reduction algorithm. Except for the decision tree, the accuracy of the other three machine learning algorithms is over 95%, and the accuracy of the decision tree algorithm is 92%. Due to the large number of parameters in the neural network algorithm, it is difficult to rewrite the algorithm, and the program does not include the neural network algorithm. For more details, please refer to Zeng et al., 2025.
It is best to use Excel versions 2019 and above for this program, as there may be compatibility issues with older versions of Excel.
A new Sphalerite Classifier.exe software has been uploaded.
Reference
Qingwen Zeng, Qihai Shu, Qingfei Wang, Qian Zhang, Zhonghai Zhao, Xudong Niu, Fan Yu, Litian Zhang, Jun Deng; Genetic types of Zn-Pb deposits revealed by sphalerite geochemistry. American Mineralogist 2025; doi: https://doi.org/10.2138/am-2024-9575