Dynamic Window-level Granger Causality of Multi-channel Time Series

Granger causality method analyzes the time series causalities without building a complex causality graph. However, the traditional Granger causality method assumes that the causalities lie between time series channels and remain constant, which cannot model the real-world time series data with dynamic causalities along the time series channels. In this paper, we present the dynamic window-level Granger causality method (DWGC) for multi-channel time series data. We build the causality model on the window-level by doing the F-test with the forecasting errors on the sliding windows. We propose the causality indexing trick in our DWGC method to reweight the original time series data. Essentially, the causality indexing is to decrease the auto-correlation and increase the cross-correlation causal effects, which improves the DWGC method. Theoretical analysis and experimental results on two synthetic and one real-world datasets show that the improved DWGC method with causality indexing better detects the window-level causalities.

[1]  Eamonn J. Keogh,et al.  On the Need for Time Series Data Mining Benchmarks: A Survey and Empirical Demonstration , 2002, Data Mining and Knowledge Discovery.

[2]  David Roxbee Cox,et al.  Nonlinear autoregressive processes , 1978, Proceedings of the Royal Society of London. A. Mathematical and Physical Sciences.

[3]  James M. Robins,et al.  Probabilistic evaluation of sequential plans from causal models with hidden variables , 1995, UAI.

[4]  Judea Pearl,et al.  Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution , 2018, WSDM.

[5]  Didier Grandjean,et al.  Time, frequency, and time‐varying Granger‐causality measures in neuroscience , 2018, Statistics in medicine.

[6]  Don H. Johnson,et al.  Statistical Signal Processing , 2009, Encyclopedia of Biometrics.

[7]  J. Nash,et al.  River flow forecasting through conceptual models part I — A discussion of principles☆ , 1970 .

[8]  Xiaohai Sun,et al.  Assessing Nonlinear Granger Causality from Multivariate Time Series , 2008, ECML/PKDD.

[9]  C. Grebogi,et al.  Inference of Granger causal time-dependent influences in noisy multivariate time series , 2012, Journal of Neuroscience Methods.

[10]  E. Fox,et al.  Neural Granger Causality for Nonlinear Time Series , 2018, 1802.05842.

[11]  N. Wermuth,et al.  Graphical Models for Associations between Variables, some of which are Qualitative and some Quantitative , 1989 .

[12]  Mukesh Dhamala,et al.  Benchmarking nonparametric Granger causality: Robustness against downsampling and influence of spectral decomposition parameters , 2018, NeuroImage.

[13]  Karl J. Friston,et al.  Dynamic causal modelling , 2003, NeuroImage.

[14]  Costas J. Spanos,et al.  CausalBG: Causal Recurrent Neural Network for the Blood Glucose Inference With IoT Platform , 2020, IEEE Internet of Things Journal.

[15]  Hao Huang,et al.  Scalable Causal Graph Learning through a Deep Neural Network , 2019, CIKM.

[16]  P. Young,et al.  Time series analysis, forecasting and control , 1972, IEEE Transactions on Automatic Control.

[17]  Rajagopalan,et al.  On the weakening relationship between the indian monsoon and ENSO , 1999, Science.

[18]  C. Granger Investigating causal relations by econometric models and cross-spectral methods , 1969 .

[19]  D. Madigan,et al.  Alternative Markov Properties for Chain Graphs , 2001 .

[20]  M. Frydenberg The chain graph Markov property , 1990 .

[21]  G. Rangarajan,et al.  Multiple Nonlinear Time Series with Extended Granger Causality , 2004 .

[22]  C. Granger Testing for causality: a personal viewpoint , 1980 .

[23]  T. Yasunari Impact of Indian monsoon on the coupled atmosphere/ocean system in the tropical pacific , 1990 .

[24]  Steven L. Scott,et al.  Inferring causal impact using Bayesian structural time-series models , 2015, 1506.00356.

[25]  Stephen Taylor,et al.  Forecasting Economic Time Series , 1979 .

[26]  Wei Liu,et al.  Discovering Granger-Causal Features from Deep Learning Networks , 2018, Australasian Conference on Artificial Intelligence.

[27]  Helmut Ltkepohl,et al.  New Introduction to Multiple Time Series Analysis , 2007 .

[28]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1971 .

[29]  M. Eichler Causal inference with multiple time series: principles and problems , 2013, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[30]  A. Rosenberg,et al.  Hume and the Problem of Causation , 1983 .

[31]  Y. Matsuda Graphical modelling for multivariate time series , 2004 .

[32]  Nicola Toschi,et al.  Echo State Network models for nonlinear Granger causality , 2019, bioRxiv.