摘要
特异值问题可能存在于许多抽样调查之中,特异值在地质勘探及采矿工程中称为特高品位,它一般对应于高品位的矿化作用.因此,特异值对于地质研究、矿石储量计算以及数据处理十分重要.本文主要讨论了以下四个问题:1)特异值及特高品位的概念;2)特异值的统计学意义;3)特异值的识别及处理方法(包括估计邻域法及影响系数法);4)用于包含特异值数据的非参数统计方法(包括指示克立格、对数正态克立格法、切尾法以及顺序一秩统计量法).
Outliers problems may be present in sampling surveys. Outliers, called extreme high grade data in geological prospecting and mining engineering, usually correspond to very high grade mineralization. So, outliers is important to geologic studing, ore reserves calculation and data processing. The paper dealt mainly with four questions:1) Concept of outliers and extreme high grade;2) statistical meaning of outliers;3) identification and handling for outliers (including: estimation neighbourhood method and influence coefficient method);4) some nonparametric statistical methods are used for data with ortliers and estimation of mean grade of blocks (including: indicator kriging, lognormal kriging, trimmed means and order - rank statistics).
出处
《地质找矿论丛》
CAS
CSCD
1990年第3期93-103,共11页
Contributions to Geology and Mineral Resources Research
