| Title | Machine learning-based delineation of geodomain boundaries: a proof-of-concept study using data from the Witwatersrand Goldfields |
| |
| Author | Zhang, S E ;
Nwaila, G T; Bourdeau, J E; Ghorbani, Y; Carranza, E J M |
| Source | Natural Resources Research vol. 32, 2023 p. 879-900, https://doi.org/10.1007/s11053-023-10159-7  Open Access |
| Image |  |
| Year | 2023 |
| Alt Series | Natural Resources Canada, Contribution Series 20230126 |
| Publisher | Springer |
| Document | serial |
| Lang. | English |
| Media | paper; digital; on-line |
| File format | pdf |
| Area | South Africa |
| Lat/Long WENS | 26.0000 28.0000 -26.0000 -28.0000 |
| Subjects | metallic minerals; Science and Technology; structural domains; gold; mineral deposits; resource estimation; metallurgical tests; Witwatersrand Basin; machine learning |
| Illustrations | location maps; tables; graphs |
| Released | 2023 03 02 |
| Abstract | Machine-aided geological interpretation provides an opportunity for rapid and data-driven decisionmaking. Indisciplines such asgeostatistics, the integrationofmachine learninghas thepotential toimprove
the reliability of mineral resourcesandore reserve estimates. Inthis study, inspiredby existing geostatistical approaches that use radial basis functions to delineate domain boundaries, we reformulate the problem into amachine learning task for
automateddomainboundarydelineation topartition theorebody.Weuse an actual dataset fromanoperatingmine(Driefontein goldmine,WitwatersrandBasininSouthAfrica)to showcase our new method.Using various machine learning algorithms, domain boundaries were
created. We show that based on a combination of in-discipline requirements and heuristic reasoning, some algorithms/ modelsmaybemoredesirable thanothers,beyondmerelycross-validationperformancemetrics. In particular, the support vector machine
algorithm yielded simple (low boundary complexity) but geologically realistic and feasible domain boundaries. In addition to the empirical results, the support vector machine algorithmis also functionally themost resemblant of current approaches that
makes use of radial basis functions. The delineated domains were subsequently used to demonstrate the effectiveness of domain delineationbycomparingdomain-based estimation versusnon-domain-based estimation using an identical automated workflow.
Analysis of estimation results indicate that domain-based estimation is more likely to result in better metal reconciliation as compared with non-domained based estimation. Through the adoption of themachine learning framework, we realized several
benefits including: uncertainty quantification; domain boundary complexity tuning; automation; dynamic updates of models using new data; and simple integration with existing machine learning-based workflows. |
| Summary | (Plain Language Summary, not published)
Inspired by existing geostatistical approaches that use radial basis functions to delineate domain boundaries, we reformulate the problem into a machine
learning task for automated domain boundary delineation to partition an orebody. |
| GEOSCAN ID | 332002 |
|
|