论文部分内容阅读
提出了一种新的矿产资源靶区定位预测的统计方法—秩特征分析方法。该方法以地质变量之间秩相关分析为基础，根据地质变量集合中某一地质变量与其余地质变量之间总的秩相关程度来度量该变量的重要性大小，根据每个地质变量在统计单元（万能的资源靶区）上的取值情况计算单元成矿联系度，，再根据单元成矿联系度相对大小评价优选矿产资源靶区。该统计方法可以同时使用定性、定量和半定量三种地质变量，减少了由于数据离散化而造成的地质信息丢失，可以最大限度地利用各种类型地质变量所提供的有用信息。
Proposed a new statistical method of target location prediction of mineral resources - rank feature analysis. Based on the rank correlation analysis between geological variables, this method measures the importance of this variable according to the total rank correlation between a certain geological variable and the remaining geological variables in the set of geological variables. Based on each geological variable in the statistical unit (Omnipotent resource target area) to calculate the unit metallogenic contact degree, and then to evaluate the preferred mineral resource target area according to the relative size of unit metallogenic contact degree. The statistical method can simultaneously use three kinds of geological variables, qualitative, quantitative and semi-quantitative, which can reduce the loss of geological information caused by data discretization and can make full use of the useful information provided by various types of geological variables.