论文信息 - Conditions for handling confounding variables in dynamic networks

Conditions for handling confounding variables in dynamic networks

Abstract In this paper we focus on consistently identifying a transfer function (module) embedded in a dynamic network. When identifying a module embedded in a dynamic network, a critical choice is which variables to include as predictor inputs. In the system identification literature sufficient conditions have been derived. One condition is that there should be no confounding variables. We show that this condition can be relaxed.

[1] Aleksandar Haber,et al. Moving Horizon Estimation for Large-Scale Interconnected Systems , 2013, IEEE Transactions on Automatic Control.

[2] Lennart Ljung,et al. System Identification: Theory for the User , 1987 .

[3] Murti V. Salapaka,et al. Identification of network components in presence of unobserved nodes , 2015, 2015 54th IEEE Conference on Decision and Control (CDC).

[4] J. Pearl. Causal inference in statistics: An overview , 2009 .

[5] Sean C. Warnick,et al. Necessary and Sufficient Conditions for Dynamical Structure Reconstruction of LTI Networks , 2008, IEEE Transactions on Automatic Control.

[6] Donatello Materassi,et al. Topological identification in networks of dynamical systems , 2008, 2008 47th IEEE Conference on Decision and Control.

[7] J. Pearl. Causality: Models, Reasoning and Inference , 2000 .

[8] Xavier Bombois,et al. Identification of dynamic models in complex networks with prediction error methods - Basic methods for consistent module estimates , 2013, Autom..

[9] Xavier Bombois,et al. Identification of Dynamic Models in Complex Networks With Prediction Error Methods: Predictor Input Selection , 2016, IEEE Transactions on Automatic Control.