格兰杰因果关系在复杂网络中的应用

An application of Granger causality in complex networks

格兰杰因果关系在经济学和生物学领域已有广泛的应用,其在计算过程中要求变量的个数远远小于时间序列的长度.为了解决实际应用中"维数灾难"的难题,把格兰杰因果关系法应用到复杂网络中.首先,利用两变量格兰杰因果关系、偏相关格兰杰因果关系,提出用迭代法一步步去除所有间接连接的步骤,从而确定每个节点的父节点,构造出复杂的网络结构;然后,再通过一个模拟的例子验证该方法的有效性.结果表明:该方法能有效地分析数据之间的内在联系.

Granger causality had been widely applied in the fields of economics and biology, though it re-quired the number of variables to be much smaller than the length of the time series in the calculation process. In order to solve the problem of the ＂curse of dimensionality＂ in practical application, and apply the Granger causality method to the complex networks, it was proposed an iterative method to remove all indirect connec-tions step-by-step by two variables Granger causality and partial correlation Granger causality. Then it was left to determine the parent nodes for each node and to construct the complex network structure. It was also veri-fied the effectiveness of the method through a simulated example. The results showed that the method could be used to analyze the intrinsic link between the data effectively.