经济预测方法选择专家系统的设计与开发

预测方法是建立预测模型的基础.因此,如何根据特定的预测对象选择适当的预测方法,是经济预测工作获得成功的关键.本文提出用知识工程方法解决经济预测方法选择的一条途径.为此,本文全面地总结了影响预测方法选择的各种因素,综合分析了这些因素与15种经济预测方法的关系,建立了知识推理网络.经济预测方法选择专家系统(简称EFM)的知识库由规则集和事实构成,共121条规则;推理机采用反向、不精确推理控制策略.该系统的开发可大大提高预测方法选择的准确性和快速性.

DESIGNING AND DEVELOPING EXPERT SYSTEM FOR SELECTING ECONOMIC FORECASTING METHOD

Forecasting methods are the fundamentals for building forecasting models. The key which makes the economic forecasting successful is how to select suitable forecasting methods in terms of the circumstances which a given forecast object is in. In this article, a new approach is put forth to solve the problem of selecting economic forecasting method using the method of knowledge engineering. For this reason, comprehensively various factors influencing the selection of forecasting method are summarized, and the relation betweeen these factors and fifteen economical forecasting methods is analyzed. In addition, knowledge reasoning network is built. The knowledge base of expert system for selecting economic forecasting method contains rule set and facts, comprising of 121 rules in the rule set. Reversed no - precising reasoning strategy is used. The development of this system will make the selection of forecasting method more accurate and speedy.