连续证据权重法及其在矿产潜力预测中的应用研究

A Continuous Fuzzy Weight-of-Evidence Model for Mineral Potential Mapping

目前分类证据权重法只能处理分类证据因子,连续证据因子在转化为分类数据时,必然导致信息损失;而且证据因子按分类计算权重,不能从整体上反映该证据因子对成矿有利的权重。该文尝试对分类证据权重法进行改进,将分类证据隶属度扩充为连续证据隶属度,并修改证据因子权重的计算方法,避免同一证据因子权重的重复累加,建立基于连续证据因子的模糊证据权重法。以实际的矿产潜力预测为例,对比分析分类证据权重法与所提出的模糊连续证据权重法。结果表明,基于连续证据因子的模糊证据权重法能够克服分类证据模型后验概率空间突变情况,利于预测结果的制图输出,并在一定程度上提高后验概率精度。

The traditional fuzzy weigh-of-evidence model can only handle categorical evidences.The categorical evidences based model preserves two main flaws.The first one is information loss when the continuous data is transformed to categorical evidences.The second one is the categorical weights can not reflect the whole effects of the evidence on the result.To overcome the shortcoming of the traditional fuzzy weigh-of-evidence model that can only handle categorial evidences,in this paper,a continuous fuzzy weight-of-evidence model for evidences with continuous attributes is proposed.The proposed model is tested in an experimental area with three continuous evidential layers.The results suggest,comparing with the traditional categorical models,that the proposed model can achieve a much higher accuracy of posterior probability and a better mapping effects for continuous mineral potential.