Quantifying Dependencies in Computer-Based Mining Simulation Modeling

Monte ? Carlo simulation is often used to assess the risks associated with uncertain variables in mining operations. The success of Monte-Carlo simulations depends upon reproducing the dependencies among variables to some extent. However, some variables may exhibit different dependency patterns. In these cases, classical techniques for dependency quantification cannot capture dependencies. As an alternative, the copulas allow modeling nonlinear dependencies. Furthermore, any choice of marginal distributions can be used. Especially, for the data, which is sensitive to extreme case such as financial crisis, the copulas arise as a means to reproduce the correlations among variables. In this paper, the copulas are introduced to mining practitioners using Monte-Carlo simulations for risk assessment. Then, a case study is given to show the superiority of the copulas over classical techniques using ModelRisk Software. In conclusion, it was observed that the copulas enhance the quality of Monte-Carlo simulations by capturing variable dependencies.