Nonlinear Causal Link Estimation Under Hidden Confounding with an Application to Time Series Anomaly Detection

Causality analysis represents one of the most important tasks when examining dynamical systems such as ecological time series. We propose to mitigate the problem of inferring nonlinear cause-effect dependencies in the presence of a hidden confounder by using deep learning with domain knowledge integration. Moreover, we suggest a time series anomaly detection approach using causal link intensity increase as an indicator of the anomaly. Our proposed method is based on the Causal Effect Variational Autoencoder (CEVAE) which we extend and apply to anomaly detection in time series. We evaluate our method on synthetic data having properties of ecological time series and compare to the vector autoregressive Granger causality (VAR-GC) baseline.

[1]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[2]  C. Granger Investigating Causal Relations by Econometric Models and Cross-Spectral Methods , 1969 .

[3]  M. Eichler Causal inference in time series analysis , 2012 .

[4]  K. Kleinman,et al.  A comparison of confounding adjustment methods with an application to early life determinants of childhood obesity. , 2014, Journal of developmental origins of health and disease.

[5]  Jonathan F. Donges,et al.  Using Causal Effect Networks to Analyze Different Arctic Drivers of Midlatitude Winter Circulation , 2016 .

[6]  Diederik P. Kingma,et al.  Variational Recurrent Auto-Encoders , 2014, ICLR.

[7]  Joachim Denzler,et al.  Deep learning and process understanding for data-driven Earth system science , 2019, Nature.

[8]  Markus Reichstein,et al.  CAUSALITY ANALYSIS OF ECOLOGICAL TIME SERIES: A TIME-FREQUENCY APPROACH , 2018 .

[9]  Yan Liu,et al.  Granger Causality for Time-Series Anomaly Detection , 2012, 2012 IEEE 12th International Conference on Data Mining.

[10]  Joachim Denzler,et al.  Detecting Regions of Maximal Divergence for Spatio-Temporal Anomaly Detection , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Diego Macias,et al.  On the causal structure between CO2 and global temperature , 2016, Scientific Reports.

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

[13]  Sepp Hochreiter,et al.  Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs) , 2015, ICLR.

[14]  João Pedro Cardoso Pereira,et al.  Unsupervised anomaly detection in time series data using deep learning , 2018 .

[15]  Max Welling,et al.  Causal Effect Inference with Deep Latent-Variable Models , 2017, NIPS 2017.

[16]  Uri Shalit,et al.  Estimating individual treatment effect: generalization bounds and algorithms , 2016, ICML.

[17]  Illtyd Trethowan Causality , 1938 .

[18]  E. H. Simpson,et al.  The Interpretation of Interaction in Contingency Tables , 1951 .

[19]  J Runge,et al.  Causal network reconstruction from time series: From theoretical assumptions to practical estimation. , 2018, Chaos.

[20]  Z. Geng,et al.  Identifying Causal Effects With Proxy Variables of an Unmeasured Confounder. , 2016, Biometrika.

[21]  J. Geweke,et al.  Measurement of Linear Dependence and Feedback between Multiple Time Series , 1982 .

[22]  Joachim Denzler,et al.  Time-Frequency Causal Inference Uncovers Anomalous Events in Environmental Systems , 2019, GCPR.

[23]  Joachim Denzler,et al.  Proceedings of the 8th International Workshop on Climate Informatics: CI 2018 , 2018, CI2018.

[24]  R. Zemel,et al.  Neural Relational Inference for Interacting Systems , 2018, ICML.

[25]  Max Welling,et al.  Auto-Encoding Variational Bayes , 2013, ICLR.

[26]  Cédric Villani,et al.  The Wasserstein distances , 2009 .