Predicting ore content throughout a machine learning procedure – An Sn-W enrichment case study

Abstract

The distribution patterns of trace elements are a very useful tool for the prediction of mineral deposits occurrence and possible future exploitation. Machine learning techniques were used for the computation of adequate models in trace elements’ prediction. The main subject of this research is to define an adequate model to predict the amounts of Sn and W in the abandoned mine area of Lardosa (Central Portugal). The geochemical composition of 333 stream sediment samples collected in the study area was used. Total concentrations of As, B, Be, Cd, Co, Cr, Cu, Fe, Ni, P, Sn, U, V, W, Y, and Zn were used to define the best prediction model. Different machine learning techniques were tested: decision trees (CART), multilayer perceptron (MLP) and support vector machines (SVM) For regression and clustering, CART, MLP and SVM for the classification problem. These algorithms were used with six different inputs – N1 to N6 - to pick out the best-performing model. The results indicate that CART achieves the best predictions for Sn and W. In the regression process, correlation coefficients of 0.67 for Sn (with Input N1) and 0.70 28 for W (with Input N3) were obtained. Regarding the classification problem, an error rate of 0.10 for both Sn (Input N1) and W (Input N2) was reached. The classification process is the best methodology to use in the prediction of Sn and W using the trace element concentration of stream sediments from Lardosa area.