上海大学学报(自然科学版) ›› 2012, Vol. 18 ›› Issue (4): 384-389.doi: 10.3969/j.issn.1007-2861.2012.04.011

• 数理化科学 • 上一篇    下一篇

基于经验模态分解的上证综合指数时间序列分析

蔡赟姝,卢志明   

  1. (上海大学 上海市应用数学和力学研究所,上海 200072)
  • 出版日期:2012-08-30 发布日期:2012-08-30
  • 通讯作者: 卢志明(1969~),男,研究员,博士,研究方向为湍流、环境流体力学. E-mail:zmlu@shu.edu.cn
  • 基金资助:

    上海大学创新基金资助项目(SHUCX112358)

Analysis of Shanghai Composite Index Time Series Based on Empirical Mode Decomposition

CAI Yun-shu,LU Zhi-ming   

  1. (Shanghai Institute of Applied Mathematics and Mechanics, Shanghai University, Shanghai 200072, China)
  • Online:2012-08-30 Published:2012-08-30

摘要: 采用经验模态分解(empirical mode decomposition, EMD)方法对上证综合指数(Shanghai composite index, SCI)进行研究,将其分解为多个内模函数(intrinsic mode functions, IMFs)和剩余项之和.通过对各阶内模函数进行基本统计分析和分布拟合,发现其“尖峰厚尾”的特点基本服从自由度为3的t分布.通过对各阶内模函数进行周期性分析,揭示各阶模态间不同的波动信息,并得到周、月、半年等时间尺度股指的波动特点,以及典型上涨和下跌时段的波动周期和波动特点.

关键词: 金融时间序列, 经验模态分解, 内模函数

Abstract: Shanghai composite index (SCI) is decomposed into a series of intrinsic mode functions (IMFs) and a residual using the empirical mode decomposition (EMD) method. Fluctuation properties of each IMF are obtained. A basic statistic analysis and fitting on the distribution of the IMFs reveal that the “leptokurtosis, fattailed” characteristics of IMFs obey tdistribution with a degree of freedom 3. Fluctuation cycles and characteristics on widely concerned time scales, such as week, month, half year, etc are obtained. Particularly, fluctuation properties of typical rising and dropping periods are analyzed. 

Key words: empirical mode decomposition (EMD), financial time series, intrinsic mode function (IMF)

中图分类号: