Mineral resource classification based on uncertainty measures in geological domains

Mineral resource classification is of paramount importance for mining industry. The main challenge for this, however, is related to the geostatistical modeling approach, in which there is no unique algorithm for such a significant act. The deterministic approaches such as kriging, indeed is not proper, because of its smoothing effect and ignoring the proportional effect that lead to possible misinterpretation of kriging variance. As an alternative, stochastic simulation based on modeling the continuous variable can be employed. Besides of legitimate criticism against this approach, it is still usable for mineral resource classification. One of the dispute is related to setting parameters and choosing the optimum Gaussian simulation algorithm. In this study, an alternative is proposed in reliance on stochastic modeling of categorical variables rather than continuous variables such as estimation domains and rock types. The algorithm is founded on probability assumption, in which definition of thresholds for different categories can be manipulated with reference to opinion of the competent person as defined in JORC code.