Robust Identification of Graph Structure

Vasileios Georgios Karanikolas; Georgios B. Giannakis

doi:10.1109/CAMSAP58249.2023.10403439

Robust Identification of Graph Structure

Vasileios Georgios Karanikolas, Georgios B. Giannakis

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Partial correlations (PCs) and the related inverse covariance matrix adopted by graphical lasso, are widely applicable tools for learning graph connectivity given nodal observations. The resultant estimators however, can be sensitive to outliers. Robust approaches developed so far to cope with outliers do not (explicitly) account for nonlinear interactions possibly present among nodal processes. This can hurt the identification of graph connectivity, merely due to model mismatch. To overcome this limitation, a novel formulation of robust PC is introduced based on nonlinear kernel functions. The proposed scheme leverages robust ridge regression techniques, spectral Fourier feature based kernel approximants, and robust association measures. Numerical tests on synthetic and real data illustrate the potential of the novel approach.

Original language	English (US)
Title of host publication	2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	486-490
Number of pages	5
ISBN (Electronic)	9798350344523
DOIs	https://doi.org/10.1109/CAMSAP58249.2023.10403439
State	Published - 2023
Event	9th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023 - Herradura, Costa Rica Duration: Dec 10 2023 → Dec 13 2023

Publication series

Name	2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023

Conference

Conference	9th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023
Country/Territory	Costa Rica
City	Herradura
Period	12/10/23 → 12/13/23

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Keywords

kernel-based methods
network topology inference
Robust statistics

Publisher link

10.1109/CAMSAP58249.2023.10403439

Find It

Find It at U of M Libraries

Cite this

Karanikolas, V. G., & Giannakis, G. B. (2023). Robust Identification of Graph Structure. In 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023 (pp. 486-490). (2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CAMSAP58249.2023.10403439

Robust Identification of Graph Structure. / Karanikolas, Vasileios Georgios; Giannakis, Georgios B.
2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 486-490 (2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Karanikolas, VG & Giannakis, GB 2023, Robust Identification of Graph Structure. in 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023. 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023, Institute of Electrical and Electronics Engineers Inc., pp. 486-490, 9th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023, Herradura, Costa Rica, 12/10/23. https://doi.org/10.1109/CAMSAP58249.2023.10403439

Karanikolas VG, Giannakis GB. Robust Identification of Graph Structure. In 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 486-490. (2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023). doi: 10.1109/CAMSAP58249.2023.10403439

Karanikolas, Vasileios Georgios ; Giannakis, Georgios B. / Robust Identification of Graph Structure. 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 486-490 (2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023).

@inproceedings{e51889a7e73b4d098c158faadee4481b,

title = "Robust Identification of Graph Structure",

abstract = "Partial correlations (PCs) and the related inverse covariance matrix adopted by graphical lasso, are widely applicable tools for learning graph connectivity given nodal observations. The resultant estimators however, can be sensitive to outliers. Robust approaches developed so far to cope with outliers do not (explicitly) account for nonlinear interactions possibly present among nodal processes. This can hurt the identification of graph connectivity, merely due to model mismatch. To overcome this limitation, a novel formulation of robust PC is introduced based on nonlinear kernel functions. The proposed scheme leverages robust ridge regression techniques, spectral Fourier feature based kernel approximants, and robust association measures. Numerical tests on synthetic and real data illustrate the potential of the novel approach.",

keywords = "kernel-based methods, network topology inference, Robust statistics",

author = "Karanikolas, {Vasileios Georgios} and Giannakis, {Georgios B.}",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 9th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023 ; Conference date: 10-12-2023 Through 13-12-2023",

year = "2023",

doi = "10.1109/CAMSAP58249.2023.10403439",

language = "English (US)",

series = "2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "486--490",

booktitle = "2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023",

}

TY - GEN

T1 - Robust Identification of Graph Structure

AU - Karanikolas, Vasileios Georgios

AU - Giannakis, Georgios B.

PY - 2023

Y1 - 2023

N2 - Partial correlations (PCs) and the related inverse covariance matrix adopted by graphical lasso, are widely applicable tools for learning graph connectivity given nodal observations. The resultant estimators however, can be sensitive to outliers. Robust approaches developed so far to cope with outliers do not (explicitly) account for nonlinear interactions possibly present among nodal processes. This can hurt the identification of graph connectivity, merely due to model mismatch. To overcome this limitation, a novel formulation of robust PC is introduced based on nonlinear kernel functions. The proposed scheme leverages robust ridge regression techniques, spectral Fourier feature based kernel approximants, and robust association measures. Numerical tests on synthetic and real data illustrate the potential of the novel approach.

AB - Partial correlations (PCs) and the related inverse covariance matrix adopted by graphical lasso, are widely applicable tools for learning graph connectivity given nodal observations. The resultant estimators however, can be sensitive to outliers. Robust approaches developed so far to cope with outliers do not (explicitly) account for nonlinear interactions possibly present among nodal processes. This can hurt the identification of graph connectivity, merely due to model mismatch. To overcome this limitation, a novel formulation of robust PC is introduced based on nonlinear kernel functions. The proposed scheme leverages robust ridge regression techniques, spectral Fourier feature based kernel approximants, and robust association measures. Numerical tests on synthetic and real data illustrate the potential of the novel approach.

KW - kernel-based methods

KW - network topology inference

KW - Robust statistics

UR - http://www.scopus.com/inward/record.url?scp=85185006623&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85185006623&partnerID=8YFLogxK

U2 - 10.1109/CAMSAP58249.2023.10403439

DO - 10.1109/CAMSAP58249.2023.10403439

M3 - Conference contribution

AN - SCOPUS:85185006623

T3 - 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023

SP - 486

EP - 490

BT - 2023 IEEE 9th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 9th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2023

Y2 - 10 December 2023 through 13 December 2023

ER -

Robust Identification of Graph Structure

Abstract

Publication series

Conference

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this