Tiger Fan, BSc
T. Fan1, H. N. Gharti1, G. Olivo1, A. Guatame-Garcia1,2
1Department of Geological Sciences and Geological Engineering, Queen’s University, Kingston, Ontario, Canada
2Department of Earth, Energy, and Environment, University of Calgary, Calgary, Alberta, Canada
As mineral exploration embraces the era of big data, the application of Artificial Intelligence (AI) techniques to analyze complex mineral systems and optimize exploration strategies becomes increasingly crucial. This work introduces AIMinex, a novel Open-Source AI-driven graphical user interface (GUI), and demonstrates its application to zinc exploration. AIMinex is designed to streamline mineral exploration, geophysical and geochemical data analysis and visualization. Leveraging advanced machine learning techniques, AIMinex enhances the analysis of complex mineralogical, geochemical, geophysical, and geological datasets, enabling more accurate mineral deposit identification and resource estimation.
AIMinex integrates a suite of tools for data visualization (including 2D and 3D plots, unsorted/sorted bar plots, and cluster plots), statistical modelling (such as Principal Component Analysis and clustering methods), and predictive analytics (featuring Random Forest and Support Vector Machine). These features allow users to interactively explore various data, uncover patterns, and make informed decisions. AIMinex is cross-platform, compatible with Windows, Linux, and Mac systems and will be publicly available to geoscientists and exploration professionals, providing a powerful yet user-friendly tool to accelerate and optimize the exploration process.
A case study involving lithogeochemical data of barren and mineralized zinc zones hosted in high-grade marbles illustrates some of the applications of AIMinex. Using Principal Component Analysis (PCA), AIMinex highlights the signature of the mineralized marbles (Zn associated with Hg, Cd, In, Mn, Ag, Cu, S and Se) compared to the non-mineralized marbles (with higher Ca and Si loadings). PCA also helps differentiate the two major styles of mineralization: one associated with dolomitic marble with traces of apatite and the other characterized by more abundant apatite (> 2%), along with a subset of mineralized samples associated with Ba, Sr, and K. Moreover, PCA reveals that lead is not directly correlated with zinc mineralized zones but occurs in association with Sb and Bi. These findings bring new insights into the understanding of zinc mineralization hosted in high-grade metamorphic terranes, showcasing the power and versatility of AIMinex for optimizing mineral exploration workflows.