摘要(Abstract)：
为了提高现有方法挖掘随时间变化的因果关系的准确性，针对包含一个因果关系区域的二元时间序列，提出通过辨识不同区域结构及结构边缘挖掘时间序列因果关系转换点的方法。该方法将因果关系区域在时间序列中的可能位置设计为左区、右区、中区结构，采用现有的因果关系发现方法探测粗略的因果关系区域，并区分为某种区域结构；根据不同区域结构的特点设计对应的边缘辨识措施，设置渐增的探测窗口及其因果强度指标，以辨识出区域结构边缘作为因果关系转换点，提高因果关系转换点的识别精度；分别在2个模拟数据集和2个真实数据集中实验验证所提方法识别因果关系转换点的准确性。结果表明，所提方法在可分离模拟数据集上使用Granger因果分数得到的因果关系转换点的平均准确性高于对比方法的，在弱耦合模拟数据集上使用收敛交叉映射因果分数得到的因果关系转换点的平均准确性在耦合程度为0.01和0.50时高于对比方法的，在2个真实数据集上使用Granger因果分数得到的因果关系转换点的准确性高于对比方法的。
关键词(KeyWords)：时间序列;因果关系;因果关系转换;收敛交叉映射;Granger因果检验
基金项目(Foundation):国家自然科学基金项目(62106049);; 福建省自然科学基金项目(2022J01656)
作者(Author): 谢杰,王开军,方莹,罗天健
DOI: 10.13349/j.cnki.jdxbn.20250226.001
参考文献(References)：
[1] ASSAAD C K,EDVIJVER E,GAUSSIER E.Survey and evaluation of causal discovery methods for time series[J].Journal of Artificial Intelligence Research,2022,73:767.
[2] DAS P,BABADI B.Non-asymptotic guarantees for reliable identification of Granger causality via the LASSO[J].IEEE Transactions on Information Theory,2023,69(11):7439.
[3] GAO W,YANG H Z.Time-varying group LASSO Granger causality graph for high dimensional dynamic system[J].Pattern Recognition,2022,130:108789.
[4] SUGIHARA G,MAY R,YE H,et al.Detecting causality in complex ecosystems[J].Science,2012,338:496.
[5] FINKLE J D,WU J J,BAGHERI N.Windowed Granger causal inference strategy improves discovery of gene regulatory networks[J].Proceedings of the National Academy of Sciences of the United States of America,2018,115(9):2252.
[6] CHANG T,TSAI S L,HAGA K A.Uncovering the interrelationship between the U.S.stock and housing markets:a bootstrap rolling window Granger causality approach[J].Applied Economics,2017,49:5841.
[7] MASNADI-SHIRAZI M,MAURYA M R,PAO G,et al.Time vary-ing causal network reconstruction of a mouse cell cycle[J].BMC Bioinformatics,2019,20:294.
[8] LI Z H,ZHENG G J,AGARWAL A,et al.Discovery of causal time intervals[C]//The 17th SIAM International Conference on Data Mining,April 27-29,2017,Houston,USA.Philadelphia:SIAM,2017:804.
[9] 王开军，曾元鹏，缪忠剑.差异区域平衡法探索时间序列变化的因果关系[J].电子与信息学报，2021,43(8):2414.
[10] GE X L,LIN A J.Kernel change point detection based on convergent cross mapping[J].Communications in Nonlinear Science and Numerical Simulation,2022,109:106318.
[11] SHI S P,HURN S,PHILLIPS P C B.Causal change detection in possibly integrated systems:revisiting the money-income relationship[J].Journal of Financial Econometrics,2020,18(1):158.
[12] TRUONG C,OUDRE L,VAYATIS N.Selective review of offline change point detection methods[J].Signal Processing,2020,167:107299.
