Extended-AUDI method for simultaneous determination of causality and models from process data
17 Jun 2013-pp 2491-2496
TL;DR: The EAUDI method is further extended to detect causality from process data, and it can also provide models of all connecting paths simultaneously and hypothesis testing is proposed to verify the results of this approach (by testing cross-regressive coefficients).
Abstract: To the best of our knowledge, there are few methods which can determine both causality and models from process data, although both of them are crucial in practical applications The extended augmented UD identification (EAUDI) is an identification approach which does not need a priori causal relationship between variables in advance In this method, however, the information contained in the augmented information matrix (AIM) is still not fully utilized and yet helpful for causality analysis, namely, whether the values of cross-regressive coefficients are sufficiently weak to be considered as insignificant Based on this, the EAUDI method is further extended to detect causality from process data, and it can also provide models of all connecting paths simultaneously Moreover, hypothesis testing (F-distribution) is proposed to verify the results of this approach (by testing cross-regressive coefficients) The effectiveness of the proposed method is demonstrated by numerical examples
Citations
More filters
01 Jan 2014
TL;DR: This chapter focuses on the relationship between different time series to capture causality in the process, and on the results of pairwise causality analysis, which can construct a causal network that is composed of the links between every two nodes.
Abstract: Data is a valuable resource for modeling and analysis. Process data is a set of timeseries of process variables. In this chapter, we focus on the relationship between different time series to capture causality in the process. For a pair of process variables, various data-based methods can be applied to detect causality. These methods can be categorized into three classes: lag-basedmethods, such as the Granger causality and transfer entropy; conditional independence methods, such as the Bayesian network; and higher order statistics, such as the Patel’s pairwise conditional probability approach. In this work, we focus on the first group of methods, which are the most commonly used, and then briefly discuss some remaining methods. Based on the results of pairwise causality analysis, one can construct a causal network that is composed of the links between every two nodes. For multivariate systems, network topology can be determined by using statistical confounding analysis.
5 citations
01 Jan 2014
TL;DR: A direct application by establishing connectivity and causality is to build a topological model before parameter identification for complex industrial processes that areusually multi-input, multi-output systems with many internal closed loops.
Abstract: Connectivity and causality have a lot of potential applications, among which we focus on analysis and design of large-scale complex industrial processes. A direct application by establishing connectivity and causality is to build a topological model before parameter identification for complex industrial processes that areusually multi-input, multi-output systems with many internal closed loops. In abnormal situation management, process topology can be employed for root cause analysis, risk analysis, and consequential alarm identification using the information of fault propagation. These potential applications include both off-line analysis andon-line diagnosis. In addition, process topology can eventually be used in design of control structures because process topology determinesthe natural structure of the distributed plant-wide control.
2 citations
TL;DR: In this paper, a generalized instrumental variable (GIV) identification method based on a UD factorization is proposed for closed-loop systems with colored noise, where all the parameter estimates and the corresponding loss function values for both forward and backward path models with orders possibly from zero to n can be obtained simultaneously after a single step of UD factorisation.
1 citations
Additional excerpts
...Property 1 (Niu et al., 1994; Niu and Fisher, 1994; Jiang et al., 2013)....
[...]
References
More filters
TL;DR: In this article, the cross spectrum between two variables can be decomposed into two parts, each relating to a single causal arm of a feedback situation, and measures of causal lag and causal strength can then be constructed.
Abstract: There occurs on some occasions a difficulty in deciding the direction of causality between two related variables and also whether or not feedback is occurring. Testable definitions of causality and feedback are proposed and illustrated by use of simple two-variable models. The important problem of apparent instantaneous causality is discussed and it is suggested that the problem often arises due to slowness in recording information or because a sufficiently wide class of possible causal variables has not been used. It can be shown that the cross spectrum between two variables can be decomposed into two parts, each relating to a single causal arm of a feedback situation. Measures of causal lag and causal strength can then be constructed. A generalisation of this result with the partial cross spectrum is suggested.
16,349 citations
TL;DR: There are several methods that can give high sensitivity to network connection detection on good quality FMRI data, in particular, partial correlation, regularised inverse covariance estimation and several Bayes net methods; however, accurate estimation of connection directionality is more difficult to achieve.
1,770 citations
TL;DR: In this paper, a frequency-domain approach to describe the relationships between multivariate time series based on the decomposition of multivariate partial coherences computed from multivariate autoregressive models is introduced.
Abstract: This paper introduces a new frequency-domain approach to describe the relationships (direction of information flow) between multivariate time series based on the decomposition of multivariate partial coherences computed from multivariate autoregressive models. We discuss its application and compare its performance to other approaches to the problem of determining neural structure relations from the simultaneous measurement of neural electrophysiological signals. The new concept is shown to reflect a frequency-domain representation of the concept of Granger causality.
1,584 citations
TL;DR: Simulation experiments have shown that the estimator proposed by the paper unequivocally reveals the direction of the signal flow and is able to distinguish between direct and indirect transfer of information.
Abstract: The paper describes the method of determining direction and frequency content of the brain activity flow. The method was formulated in the framework of the AR model. The transfer function matrix was found for multichannel EEG process. Elements of this matrix, properly normalized, appeared to be good estimators of the propagation direction and spectral properties of the investigated signals. Simulation experiments have shown that the estimator proposed by us unequivocally reveals the direction of the signal flow and is able to distinguish between direct and indirect transfer of information. The method was applied to the signals recorded in the brain structures of the experimental animals and also to the human normal and epileptic EEG. The sensitivity of the method and its usefulness in the neurological and clinical applications was demonstrated.
1,000 citations
"Extended-AUDI method for simultaneo..." refers background in this paper
...Several approaches have been carried out to study the causal relationships in processes [2,3], such as process flow diagram (PFD) [4], construction of signed digraph (SDG) [5], adjacency matrix [6], Partial directed coherence (PDC) [7] and directed transfer function (DTF) [8]....
[...]
TL;DR: This work describes the multivariate linear methods most commonly used in neurophysiology and shows that they can be extended to assess the existence of nonlinear interdependence between signals and describes nonlinear methods based on the concepts of phase synchronization, generalized synchronization and event synchronization.
993 citations
"Extended-AUDI method for simultaneo..." refers background in this paper
...Several approaches have been carried out to study the causal relationships in processes [2,3], such as process flow diagram (PFD) [4], construction of signed digraph (SDG) [5], adjacency matrix [6], Partial directed coherence (PDC) [7] and directed transfer function (DTF) [8]....
[...]