Mineral resources prediction and evaluation based on an improved entropy weight method and uncertainty measure theory

The sources of uncertainty in quantitative mineral prediction and evaluation are identified and classified through systematic analysis of three data error types. In view of the inherent deficiencies of the conventional entropy weight approach, an improved formulation is introduced to resolve its limited applicability in specific cases. In this article, an improved entropy weight method combined with uncertainty measure theory is employed to construct a comprehensive model for quantitative mineral prediction and uncertainty assessment. This model is then applied to evaluate the metallogenic favorability of skarn-type polymetallic deposits across seven ore belts within the Hutouya orefield in Qinghai, China. To this end, 11 influencing factors are selected, for which linear-type uncertainty measure functions are established and factor weights are determined using the improved entropy weight method. Finally, the metallogenic favorability of the seven ore belts is ranked and evaluated based on the credible degree recognition criterion. The results show that the proposed evaluation process is both scientifically rigorous and operationally feasible. Moreover, its strong practical utility offers a valuable reference for future quantitative prediction and uncertainty assessment in mineral resource exploration.