基于粗糙集理论的铁路货运需求预测研究

Demand Forecast of Rail Freight Based on Rough Set Theory

针对我国铁路货运量的特点,较全面的选取了铁路货运需求预测指标体系;基于粗糙集理论,建立了铁路货运需求预测的知识库,分别应用等距离法、属性约简进行数据离散化和提取关键指标;对我国铁路货运量进行了比较准确的预测,并计算出了各个关键指标的重要程度。结果表明,应用粗糙集理论进行铁路货运量预测时,可以直接从指标数据本身出发,有效避免了主观因素的影响,提高了预测的可靠性和有效性,为制定提升铁路货运量的政策提供了比较合理的理论依据。

Based on characteristics of the railway freight, the paper selects the index system of demand forecast of rail freight comprehensively. Then based on rough set theory, the paper establishes a knowledge base of demand forecast of rail freight, discrete data and extracts key indicators by using the equidistance method and attribute reduction method. The paper predicts freight volume of China~ railway accurately and calculates the degree of importance of each key indica- tor. The results shows that when we predicts the rai! freight, we can start directly from the index data itself. This avoids the influence of subjective factors, improves the reliability and validity of the prediction and provides a reasonable theoret- ical basis for policies to improve rail freight.