Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models

Biwei Huang, Kun Zhang, Mingming Gong, Clark Glymour
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:2901-2910, 2019.

Abstract

In many scientific fields, such as economics and neuroscience, we are often faced with nonstationary time series, and concerned with both finding causal relations and forecasting the values of variables of interest, both of which are particularly challenging in such nonstationary environments. In this paper, we study causal discovery and forecasting for nonstationary time series. By exploiting a particular type of state-space model to represent the processes, we show that nonstationarity helps to identify the causal structure, and that forecasting naturally benefits from learned causal knowledge. Specifically, we allow changes in both causal strengths and noise variances in the nonlinear state-space models, which, interestingly, renders both the causal structure and model parameters identifiable. Given the causal model, we treat forecasting as a problem in Bayesian inference in the causal model, which exploits the time-varying property of the data and adapts to new observations in a principled manner. Experimental results on synthetic and real-world data sets demonstrate the efficacy of the proposed methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-huang19g, title = {Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models}, author = {Huang, Biwei and Zhang, Kun and Gong, Mingming and Glymour, Clark}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {2901--2910}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/huang19g/huang19g.pdf}, url = {https://proceedings.mlr.press/v97/huang19g.html}, abstract = {In many scientific fields, such as economics and neuroscience, we are often faced with nonstationary time series, and concerned with both finding causal relations and forecasting the values of variables of interest, both of which are particularly challenging in such nonstationary environments. In this paper, we study causal discovery and forecasting for nonstationary time series. By exploiting a particular type of state-space model to represent the processes, we show that nonstationarity helps to identify the causal structure, and that forecasting naturally benefits from learned causal knowledge. Specifically, we allow changes in both causal strengths and noise variances in the nonlinear state-space models, which, interestingly, renders both the causal structure and model parameters identifiable. Given the causal model, we treat forecasting as a problem in Bayesian inference in the causal model, which exploits the time-varying property of the data and adapts to new observations in a principled manner. Experimental results on synthetic and real-world data sets demonstrate the efficacy of the proposed methods.} }
Endnote
%0 Conference Paper %T Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models %A Biwei Huang %A Kun Zhang %A Mingming Gong %A Clark Glymour %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-huang19g %I PMLR %P 2901--2910 %U https://proceedings.mlr.press/v97/huang19g.html %V 97 %X In many scientific fields, such as economics and neuroscience, we are often faced with nonstationary time series, and concerned with both finding causal relations and forecasting the values of variables of interest, both of which are particularly challenging in such nonstationary environments. In this paper, we study causal discovery and forecasting for nonstationary time series. By exploiting a particular type of state-space model to represent the processes, we show that nonstationarity helps to identify the causal structure, and that forecasting naturally benefits from learned causal knowledge. Specifically, we allow changes in both causal strengths and noise variances in the nonlinear state-space models, which, interestingly, renders both the causal structure and model parameters identifiable. Given the causal model, we treat forecasting as a problem in Bayesian inference in the causal model, which exploits the time-varying property of the data and adapts to new observations in a principled manner. Experimental results on synthetic and real-world data sets demonstrate the efficacy of the proposed methods.
APA
Huang, B., Zhang, K., Gong, M. & Glymour, C.. (2019). Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:2901-2910 Available from https://proceedings.mlr.press/v97/huang19g.html.

Related Material