【作者】 张颖； 路影；

【Author】 ZHANG Ying LU Ying (School of Systems Engineering Graduate,Shenyang University of Technology,Shenyang 110023,China.

【机构】 沈阳工业大学 系统工程研究所；

【摘要】 在研究时尚商品特殊属性和价格特点的基础上,分析了适合进行多重分形分析的时尚商品类型,针对大众时尚商品的价格时间序列,采用 G-P 算法计算出时尚商品多重分形分布的广义维数谱,验证了该类时尚商品价格时间序列的多重分形特征．在分析样本数据特点基础上,提出用多重分形谱的直接算法计算该类时尚商品价格时间序列的多重分形谱,并讨论了奇异指数α的变化幅度与价格变化的关系和多重分形谱参数 f(α)变化与价格变化的关系．通过实例分析,得到大众时尚商品价格时间序列多重分形谱变化与其价格变化具有一致性的变化规律,通过时间序列多重分形谱的变化,进而进行价格的预测．更多还原

【Abstract】 Based on the special attribute and price characteristic of the fashion commodity,the fashion commodity type that fits multi-fractal analysis is analyzed.Aiming at the time series of the popular fashion commodity price, Grassberger-Procaccia (G-P) algorithm is adopted to calculate the generalized dimension spectrum of the multi-fractal distribution of the fashion commodity price time series,which proves the multi-fractal features of the time series. Based on analyzing the sample characteristic,the multi-fractal spectrum is calculated by using the direct algorithm. The relationship between the variety range of oddity exponent and price variety is discussed as well as that between the variety of multi-fractal spectrum parameter and the price variety.Example analysis show that the variety of multi- fractal spectrum of the popular fashion price time series and that of its price are concurrent,and the price is forecasted by the variety of multi-fractal spectrum.更多还原