改进的加权证据权模型及其在个旧锡铜矿产资源预测中的应用

为了探讨新的加权系数估计方法对于消除或减弱证据层不满足条件独立性假设时对预测结果的影响, 对加权证据权模型的加权系数估计方法进行了新的探讨，尝试用顺序估计法估计加权系数.加权系数的顺序估计法是将加权证据权模型与基于模糊预测对象的证据权模型相结合，将证据层按照一定顺序逐步加入到加权证据权模型中，在加入到模型的过程中依次用已经获得的后验概率作为模糊训练层对证据层加入到模型的顺序进行修正，并通过条件相关系数的方法估计加权系数.分别以1组多元正态分布模拟数据和个旧锡铜多金属矿产资源预测为例，比较了多种模型的后验概率，结果表明加权证据权模型对减弱证据层不满足条件独立性假设所产生的影响是有效的. 

Improvement of Weighted Weights of Evidence and Its Applications in Sn-Cu Mineral Potential Mapping in Gejiu,Yunnan Province,China

This paper proposes a new method to estimate weighting coefficients in weighted weights of evidence (WWofE) in order to reduce the influence of correlation among evidence layers when hypothesis of conditional independence is not true. Sequential estimation method in WWofE combines WWofE and dual weights of evidence (DWofE) to gradually add evidence layers with specific sequence to WWofE model. In the process, the posterior probability obtained by former evidence layers is considered as fuzzy training layer in DWofE to correct the sequence of layers. The weighting coefficients are estimated by using conditional correlation coefficient. As a case study, a group data generated by four variable normal distribution and Sn-Cu mineral resources assessment in Gejiu, Yunan, southwestern China are used. The results show that WWofE model is effective to reduce the influence of correlations among evidence layers.