Dynamic network reconstruction from heterogeneous datasets
(2021) In Automatica 123.- Abstract
Performing multiple experiments is common when learning internal mechanisms of complex systems. These experiments can include perturbations of parameters or external disturbances. A challenging problem is to efficiently incorporate all collected data simultaneously to infer the underlying dynamic network. This paper addresses the reconstruction of dynamic networks from heterogeneous datasets under the assumption that the underlying networks share the same Boolean structure across all experiments. Parametric models are derived for dynamical structure functions, which describe causal interactions between measured variables. Multiple datasets are integrated into one regression problem with additional demands on group sparsity to assure... (More)
Performing multiple experiments is common when learning internal mechanisms of complex systems. These experiments can include perturbations of parameters or external disturbances. A challenging problem is to efficiently incorporate all collected data simultaneously to infer the underlying dynamic network. This paper addresses the reconstruction of dynamic networks from heterogeneous datasets under the assumption that the underlying networks share the same Boolean structure across all experiments. Parametric models are derived for dynamical structure functions, which describe causal interactions between measured variables. Multiple datasets are integrated into one regression problem with additional demands on group sparsity to assure network sparsity and structure consistency. To acquire structured group sparsity, we propose a sampling-based method, together with extended versions of l1-methods and sparse Bayesian learning. The performance of the proposed methods is benchmarked in numerical simulation. In summary, this paper presents efficient methods on network reconstruction from multiple experiments, and reveals practical experience that could guide applications.
(Less)
- author
- Yue, Zuogong ; Thunberg, Johan LU ; Pan, Wei ; Ljung, Lennart and Gonçalves, Jorge
- publishing date
- 2021-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Heterogeneity, Multiple experiments, Network reconstruction, Sparsity, System identification
- in
- Automatica
- volume
- 123
- article number
- 109339
- publisher
- Pergamon Press Ltd.
- external identifiers
-
- scopus:85096700788
- ISSN
- 0005-1098
- DOI
- 10.1016/j.automatica.2020.109339
- language
- English
- LU publication?
- no
- additional info
- Publisher Copyright: © 2020 Elsevier Ltd
- id
- 6d198de2-97f9-495f-9c95-930f6109fba4
- date added to LUP
- 2024-09-05 09:03:04
- date last changed
- 2024-09-05 15:47:41
@article{6d198de2-97f9-495f-9c95-930f6109fba4,
abstract = {{<p>Performing multiple experiments is common when learning internal mechanisms of complex systems. These experiments can include perturbations of parameters or external disturbances. A challenging problem is to efficiently incorporate all collected data simultaneously to infer the underlying dynamic network. This paper addresses the reconstruction of dynamic networks from heterogeneous datasets under the assumption that the underlying networks share the same Boolean structure across all experiments. Parametric models are derived for dynamical structure functions, which describe causal interactions between measured variables. Multiple datasets are integrated into one regression problem with additional demands on group sparsity to assure network sparsity and structure consistency. To acquire structured group sparsity, we propose a sampling-based method, together with extended versions of l<sub>1</sub>-methods and sparse Bayesian learning. The performance of the proposed methods is benchmarked in numerical simulation. In summary, this paper presents efficient methods on network reconstruction from multiple experiments, and reveals practical experience that could guide applications.</p>}},
author = {{Yue, Zuogong and Thunberg, Johan and Pan, Wei and Ljung, Lennart and Gonçalves, Jorge}},
issn = {{0005-1098}},
keywords = {{Heterogeneity; Multiple experiments; Network reconstruction; Sparsity; System identification}},
language = {{eng}},
publisher = {{Pergamon Press Ltd.}},
series = {{Automatica}},
title = {{Dynamic network reconstruction from heterogeneous datasets}},
url = {{http://dx.doi.org/10.1016/j.automatica.2020.109339}},
doi = {{10.1016/j.automatica.2020.109339}},
volume = {{123}},
year = {{2021}},
}