Abstract:Risk forecasting is an important issue in the environment economic evaluation of mining projects. Traditional risk-forecasting methods need experts to estimate the risk probability by experience, and it is difficult to form a constantly scoring criteria. In this article, we propose a new environmental risk-forecasting method and apply it to the environmental risk prediction for economic evaluation of mining projects. Among the many variables that may affect the environmental risk, we firstly use principal component analysis (PCA) method to reduce data dimensionality. After that, for each kind of risk, we conduct a risk forecasting Via probabilistic support vector machine (PSVM). Contrasting to previous methods, this method lowers the demand for experts and it is easier to address the environmental risk-forecasting problem. Experimental results show the effectiveness of the proposed approach.

A New Risk Probability Forecasting Method For Economic Evaluation On Mining Projects