摘要
本文提出一种新的预测方法─可拓聚类预测方法，并用该方法进行邮电 业务总量预测的研究。利用邮电业务总量和三大产业的生产总值的历史 数据构造分类经典域和节域物元，应用物元和可拓集合中的关联函数， 建立了业务量水平聚类预测的一种简单模型，通过聚类分析得到邮电业 务量预测结果。本文利用我国某地区的实际数据进行分析计算，结果表 明可拓聚类预测方法是有效的。
A new forecasting method,the extension classified prediction, is proposed in this paper. The model is used to forecast the business total of posts and telecommunications, The history data are adopted to set up the classical and extensional matter-elements. The result of the business total of posts and telecommunications forecast is obtained by applying the classified anal- ysis methods. A set of real data is used,and the result is satisfied.
出处
《系统工程》
CSCD
2000年第3期73-77,共5页
Systems Engineering
关键词
邮电业务总量
可拓集合
聚类分析
预测
Business total of posts and telecommunications, Extension set,Classified analysis, Dependent function, Prediction
