摘要
本文基于粗糙集理论给出了对象与规则可信度的概念,提出了一种经济系统中相关关系预测法。利用信息熵的概念,评价因素的重要程度和约简因素,由此提取基本相关规则。在对象与规则之间的可信度基础上,建立了预测模型。与传统的回归预测法比较,这个方法不需要进行相关性判断、模型识别和检验,它直接从数据出发,在不损失信息的条件下约简冗余因素,寻找经济指标与影响因素之间的相关关系,能同时处理定性、定量因素以及不确定因素。算例计算说明了所提方法的有效性。
According to rough set theories, this paper puts forward a conception of the degree of credibility between objects and rules,and attempts to present a related relation prediction method. Information entropy is employed to reduce factors, assess the importance of them and derive basic related rules. On the basis of the degree of credibility, a prediction model is presented. Compared with traditional regression predictions, this approach does not need to proceed relativity judgment, pattern recognition and examination, while it will reduce redundancy factors without losing information directly from data in order to mine the related relation between economics indices and influent factors. This method can deal with qualitative, quantitative and uncertainty factors. The results proves the usefulness of the proposed method for a practical mediumterm forecasting.
出处
《运筹与管理》
CSCD
2002年第6期20-26,共7页
Operations Research and Management Science
基金
重庆市科学技术计划资助项目(7117)
关键词
粗糙集
相关关系
信息熵
预测
经济系统
rough set
related relation
information entropy
prediction
