BWM-MARCOS: A new hybrid MCDM approach for mineral potential modelling
Year: 2024
Type: article
Abstract: Here we describe the application of a novel hybrid multi-criteria decision-making (MCDM) approach termed “Best Worst Method-Measurement of Alternatives and Ranking according to COmpromise Solution” (BWM-MARCOS) to mineral potential modelling (MPM). The BWM-MARCOS technique combines two mathematical frameworks in which the BWM approach is utilised to weight decision criteria (i.e., predictor maps) whilst the MARCOS approach is applied to rank alternatives. A BWM-MARCOS model of orogenic gold mineral systems in the well-endowed Granites-Tanami Orogen (GTO) in Australia’s Northern Territory was generated utilising the same study area and competent spatial proxies as previously developed and described by Roshanravan et al. (2020). We found the BWM-MARCOS approach to MPM performed better than any of the knowledge-driven (i.e., fuzzy inference system), continuous (i.e., data-driven index overlay, geometric average and fuzzy gamma) and data-driven (i.e., feed-forward deep neural network and ‘original’ random forest) mineral prospectivity models previously developed by Roshanravan et al. (2020, 2023). To further restrict the BWM-MARCOS modelling outputs, the gold potential zones were delimited utilising a concentration-area (C-A) fractal approach. The areas covered by the prioritised 1st order (very high priority) and 2nd order (high priority) targets represent significant reductions of the search space (i.e., >303 times or two orders of magnitude, and >93 times or one order of magnitude, respectively). Any order of magnitude, or greater, reduction of the search space can be considered a hallmark of a well-performing, practically useful targeting approach.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,730
OpenAlex ID: W4404499833