Physica A: Statistical Mechanics and its Applications ( IF 2.8 ) Pub Date : 2021-04-02 , DOI: 10.1016/j.physa.2021.125990 E.F. Guedes , A.M. da Silva Filho , G.F. Zebende
In this paper we implemented the detrended multiple cross-correlation coefficient with sliding windows approach in order to measure multiple cross-correlation between time series as a function of time scale , and a sliding time . Faced with several non-stationary time series, a fixed size window will be defined by , and the coefficients are calculated for different time scales () and time . Following this methodology, it was possible to set up a cross-correlation color map for a simulated (random) and an empirical case. This new color map with sliding windows shows the time scale and the current day for a cross-correlation analysis. In this way, it is possible to analyze the multiple cross-correlation between three or more variables in different areas of knowledge, such as economics and finance.
中文翻译:
滑动窗口法去趋势多重相关系数
在本文中,我们使用滑动窗口方法实现了去趋势的多重互相关系数,以便测量时间序列之间的多重互相关作为时间标度的函数 ,以及滑动时间 。面对几个非平稳时间序列,固定大小的窗口将由,以及系数 是针对不同的时间范围计算的()和时间 。按照这种方法,可以建立一个模拟(随机)和经验案例的互相关色图。带有滑动窗口的新色图显示了时间标度和当前日期,以进行互相关分析。这样,可以分析不同知识领域(例如,经济和金融)中三个或更多变量之间的多重互相关。