论文部分内容阅读
回顾铝土矿地质工作中应用计算机和地质数学的经验教训,指出对产地或矿床的地质情况没有正确理解就不会有正确的数学评价。只有属于同一数学总体的数据才应当一同加以评价,忽视越轨值可能会严重歪曲计算结果,数据类型和数据分布也必须仔细地考虑。迄今有关铝土矿地质的大多数计算分析属于单变量统计分析,两种主要的二元统计分析方法是相关分析和回归分析。主要的多元统计分析方法有复相关分析,多元回归分析,多元方差分析、典型相关分析;和典型群:聚合分析、非线性映象分析、辨别分析、主成分分析、因素分析、对应分析。论述各种分析方法的适用范围,和选择分析方法的工作步骤。指出计算机科学和地质数学的发展,提供了更多更广的应用可能性,包括多元统计分析、半定量数据统计分析、模糊地质统计分析、非参数统计分析、计算机化综合信息系统、专家系统。
Reviewing the experience of lessons learned in the application of computer and geological mathematics in the geological work of bauxite, it is pointed out that if there is no correct understanding of the geology of the producing area or deposit, there will be no correct mathematical evaluation. Only data that belongs to the same mathematical population should be evaluated together. Ignoring derailment values can seriously distort the calculation results. Data types and data distribution must also be carefully considered. Most of the calculations related to bauxite geology so far are univariate statistical analyses. The two main binary statistical analysis methods are correlation analysis and regression analysis. The main multivariate statistical analysis methods include complex correlation analysis, multiple regression analysis, multivariate analysis of variance, and canonical correlation analysis; and typical clusters: aggregation analysis, nonlinear mapping analysis, discrimination analysis, principal component analysis, factor analysis, and correspondence analysis. Describe the scope of application of various analytical methods, and select the analytical method of work steps. It points out that the development of computer science and geology and mathematics provides more and wider application possibilities, including multivariate statistical analysis, semi-quantitative data statistical analysis, fuzzy geological statistical analysis, non-parametric statistical analysis, computerized integrated information system, and expert system.