中长期预测模型的GMDH两水平算法的改进及实证分析比较  

Improvement Two-level Algorithm of Model of Intermediate-or Long-term Predection and Application

作　　者：田益祥[1] 陈华富[2] 

机构地区：[1]华中科技大学经济学院 [2]电子科技大学应用数学系,四川成都610054

出　　处：《系统工程理论与实践》2002年第8期73-78,共6页Systems Engineering-Theory & Practice

摘　　要：两水平算法是以自组织原理为基础应用多层迭代、采用季节数据和年数据同时建模 ,在一定程度上扩大可预测范围 .主要用于复杂系统建模 .对算法作改进 ,使模型在对具有周期趋势的事物预测更加准确 .实证分析表明 ,对算法改进有效 .Based on the principle of self\|orgnization two\|level algorithm model system with mulitilayer algorithm Modeling simultaneously with annual and seasonal data to increase the reliability of prediction. It mainly solves the problems of patten recognization complex system modeling and intermediate or long\|term prediction .In this article, we improved the algorithm make its prediction is better.

关 键 词：中长期预测模型 GMDH 实证分析 两水平算法 自组织原理 物价 

分 类 号：F714.1[经济管理—产业经济]