当前位置: X-MOL 学术Math. Popul. Stud. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A consistency algorithm based on information theory
Mathematical Population Studies ( IF 1.8 ) Pub Date : 1994-02-01 , DOI: 10.1080/08898489409525373
Evert Van Imhoff

This paper provides a geometric‐mean solution to the consistency problem of multidimensional demographic projection models, based on the constrained minimization of an entropy function. A comparison with the existing harmonic‐mean solution yields many similarities and almost no differences: both solutions satisfy the properties of availability, monotonicity, homogeneity, competition, and symmetry, and, for both solutions, there is a convenient one‐to‐one relationship between adjustments in aggregate numbers of events, on the one hand, and age‐specific numbers of events, on the other hand. However, one major advantage of the geometric mean is that its corresponding distance function is firmly based on (information) theory.

中文翻译:

基于信息论的一致性算法

本文基于熵函数的约束最小化,为多维人口统计投影模型的一致性问题提供了几何平均解决方案。与现有调和平均解的比较有许多相似之处,几乎没有区别:两种解决方案都满足可用性、单调性、同质性、竞争性和对称性的特性,并且对于这两种解决方案,存在方便的一对一关系一方面对事件总数进行调整,另一方面对特定年龄的事件数量进行调整。然而,几何平均的一大优点是其对应的距离函数牢固地基于(信息)理论。
更新日期:1994-02-01
down
wechat
bug