基于分形理论的时尚商品价格预测问题研究  

Study on fashion commodity price forecasting issues based on fractal theory

作　　者：张颖[1] 路影[1] 么艳 

机构地区：[1]沈阳工业大学系统工程研究所,沈阳110023 [2]沈阳农村经济委员会会计核算中心,沈阳110001

出　　处：《沈阳工业大学学报（社会科学版）》2008年第2期148-150,175,共4页Journal of Shenyang University of Technology(Social Sciences)

基　　金：国家自然科学基金管理科学部重点资助项目(70431003)

摘　　要：在研究时尚商品价格规律及影响因素的复杂性、不确定性的基础之上,提出了基于分形理论的时尚商品价格预测新方法。该方法首先运用重标极差法分析了时尚商品价格时间序列预测的可行性,然后根据分形统计模型,得到时尚商品历史价格时间序列的分形维数,通过不断增加新销售或预测得到新的记录方法,求得增加新记录后的新分形维数,由此可以预测出下一时间单位的商品销售价格。通过实例进行了方法检验和结果比较,取得了较为理想的预测结果,证明了该方法的可行性和有效性。Based on the study of price law of fashion commodity and the complication and uncertainty of its influencing factors,a new forecasting method based on fractal theory was put forward.Rescaled range analysis was adopted to analyze the fractal features of fashion commodity,and the fractal dimension of fashion commodity time series was obtained on the basis of the fractal statistical model.The price of this commodity in the next time unit can be forecasted by the changing fractal dimension,which is gotten by a new method adding ordinal new record continually.Finally,by certification of the method and comparison of the result,the forecasting result is proved to be ideal,which means that the method is feasible and effective.

关 键 词：时尚商品 重标极差法 价格时间序列 价格预测 分形统计模型 

分 类 号：F224[经济管理—国民经济]