• 中国中文核心期刊
  • 中国科学引文数据库(CSCD)核心库来源期刊
  • 中国科技论文统计源期刊(CJCR)
  • 第二届国家期刊奖提名奖

Citation:

The Choice of Variables Transformation and Estimate of Measurement Error Matrix for the Study on Density diameter Relationship

  • Received Date: 1995-12-12
  • It is necessary to determine the ratio between variance of sampling error e of quadratic average diameter Dg and that of sampling error ∈ of the tree numbers N per unit area when measurement error models is applied to estimate the parameters of a model between Dg and N,in a study on stand self thinning processes.This paper calculated and compared the variance ratios between transformd quadratic average diameter and transformd number of trees per unit area by using resampling from 20 larch plantation plots.The results show that the ratio between variance of ln( N ) and Dg approaches to a constant 3.0248 ×10-2. From the results,the best form of variable transformation is determined and the structure of error matrix is constructed,which can be used to get more accurate estimate of model parameters.
  • 加载中
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Article views(2768) PDF downloads(1449) Cited by()

Proportional views

The Choice of Variables Transformation and Estimate of Measurement Error Matrix for the Study on Density diameter Relationship

  • 1. The Research Institute of Forest Resources Information and Technique, CAF Beijing100091

Abstract: It is necessary to determine the ratio between variance of sampling error e of quadratic average diameter Dg and that of sampling error ∈ of the tree numbers N per unit area when measurement error models is applied to estimate the parameters of a model between Dg and N,in a study on stand self thinning processes.This paper calculated and compared the variance ratios between transformd quadratic average diameter and transformd number of trees per unit area by using resampling from 20 larch plantation plots.The results show that the ratio between variance of ln( N ) and Dg approaches to a constant 3.0248 ×10-2. From the results,the best form of variable transformation is determined and the structure of error matrix is constructed,which can be used to get more accurate estimate of model parameters.

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return