Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors

Alexander Dorndorf; Boris Kargoll; Jens-André Paffenholz; Hamza Alkhatib

doi:10.1007/1345_2023_210

Details

Original language	English
Title of host publication	X Hotine-Marussi Symposium on Mathematical Geodesy
Subtitle of host publication	Proceedings of the Symposium, 2022
Editors	Jeffrey T. Freymueller, Laura Sánchez
Place of Publication	Berlin, Heidelberg
Pages	93-99
Number of pages	7
Publication status	Published - 2024

Publication series

Name	International Association of Geodesy Symposia
Volume	155
ISSN (Print)	0939-9585
ISSN (electronic)	2197-9359

Abstract

Geodetic measurements rely on high-resolution sensors, but produce data sets with many observations which may contain outliers and correlated deviations. This paper proposes a powerful solution using Bayesian inference. The observed data is modeled as a multivariate time series with a stationary autoregressive (VAR) process and multivariate t-distribution for white noise. Bayes’ theorem integrates prior knowledge. Parameters, including functional, VAR coefficients, scaling, and degree of freedom of the t-distribution, are estimated with Markov Chain Monte Carlo using a Metropolis-within-Gibbs algorithm.

Keywords

Metropolis-within-Gibbs algorithm, Robust Bayesian time series analysis, VAR process, t-distribution

ASJC Scopus subject areas

Earth and Planetary Sciences(all)
Computers in Earth Sciences
Earth and Planetary Sciences(all)
Geophysics

Cite this

Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors. / Dorndorf, Alexander; Kargoll, Boris; Paffenholz, Jens-André et al.
X Hotine-Marussi Symposium on Mathematical Geodesy: Proceedings of the Symposium, 2022. ed. / Jeffrey T. Freymueller; Laura Sánchez. Berlin, Heidelberg, 2024. p. 93-99 (International Association of Geodesy Symposia; Vol. 155).

Research output: Chapter in book/report/conference proceeding › Contribution to book/anthology › Research › peer review

Dorndorf, A, Kargoll, B, Paffenholz, J-A & Alkhatib, H 2024, Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors. in JT Freymueller & L Sánchez (eds), X Hotine-Marussi Symposium on Mathematical Geodesy: Proceedings of the Symposium, 2022. International Association of Geodesy Symposia, vol. 155, Berlin, Heidelberg, pp. 93-99. https://doi.org/10.1007/1345_2023_210

Dorndorf, A., Kargoll, B., Paffenholz, J.-A., & Alkhatib, H. (2024). Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors. In J. T. Freymueller, & L. Sánchez (Eds.), X Hotine-Marussi Symposium on Mathematical Geodesy: Proceedings of the Symposium, 2022 (pp. 93-99). (International Association of Geodesy Symposia; Vol. 155).. https://doi.org/10.1007/1345_2023_210

Dorndorf A, Kargoll B, Paffenholz JA, Alkhatib H. Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors. In Freymueller JT, Sánchez L, editors, X Hotine-Marussi Symposium on Mathematical Geodesy: Proceedings of the Symposium, 2022. Berlin, Heidelberg. 2024. p. 93-99. (International Association of Geodesy Symposia). Epub 2023 Sept 6. doi: 10.1007/1345_2023_210

Dorndorf, Alexander ; Kargoll, Boris ; Paffenholz, Jens-André et al. / Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors. X Hotine-Marussi Symposium on Mathematical Geodesy: Proceedings of the Symposium, 2022. editor / Jeffrey T. Freymueller ; Laura Sánchez. Berlin, Heidelberg, 2024. pp. 93-99 (International Association of Geodesy Symposia).

Download

@inbook{f230383144ed4cbd899c136a230e0b45,

title = "Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors",

abstract = "Geodetic measurements rely on high-resolution sensors, but produce data sets with many observations which may contain outliers and correlated deviations. This paper proposes a powerful solution using Bayesian inference. The observed data is modeled as a multivariate time series with a stationary autoregressive (VAR) process and multivariate t-distribution for white noise. Bayes{\textquoteright} theorem integrates prior knowledge. Parameters, including functional, VAR coefficients, scaling, and degree of freedom of the t-distribution, are estimated with Markov Chain Monte Carlo using a Metropolis-within-Gibbs algorithm.",

keywords = "Metropolis-within-Gibbs algorithm, Robust Bayesian time series analysis, VAR process, t-distribution",

author = "Alexander Dorndorf and Boris Kargoll and Jens-Andr{\'e} Paffenholz and Hamza Alkhatib",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2023.",

year = "2024",

doi = "10.1007/1345_2023_210",

language = "English",

isbn = "9783031553592",

series = "International Association of Geodesy Symposia",

pages = "93--99",

editor = "Freymueller, {Jeffrey T.} and Laura S{\'a}nchez",

booktitle = "X Hotine-Marussi Symposium on Mathematical Geodesy",

}

Download

TY - CHAP

T1 - Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors

AU - Dorndorf, Alexander

AU - Kargoll, Boris

AU - Paffenholz, Jens-André

AU - Alkhatib, Hamza

N1 - Publisher Copyright: © The Author(s) 2023.

PY - 2024

Y1 - 2024

N2 - Geodetic measurements rely on high-resolution sensors, but produce data sets with many observations which may contain outliers and correlated deviations. This paper proposes a powerful solution using Bayesian inference. The observed data is modeled as a multivariate time series with a stationary autoregressive (VAR) process and multivariate t-distribution for white noise. Bayes’ theorem integrates prior knowledge. Parameters, including functional, VAR coefficients, scaling, and degree of freedom of the t-distribution, are estimated with Markov Chain Monte Carlo using a Metropolis-within-Gibbs algorithm.

AB - Geodetic measurements rely on high-resolution sensors, but produce data sets with many observations which may contain outliers and correlated deviations. This paper proposes a powerful solution using Bayesian inference. The observed data is modeled as a multivariate time series with a stationary autoregressive (VAR) process and multivariate t-distribution for white noise. Bayes’ theorem integrates prior knowledge. Parameters, including functional, VAR coefficients, scaling, and degree of freedom of the t-distribution, are estimated with Markov Chain Monte Carlo using a Metropolis-within-Gibbs algorithm.

KW - Metropolis-within-Gibbs algorithm

KW - Robust Bayesian time series analysis

KW - VAR process

KW - t-distribution

UR - http://www.scopus.com/inward/record.url?scp=85195465959&partnerID=8YFLogxK

U2 - 10.1007/1345_2023_210

DO - 10.1007/1345_2023_210

M3 - Contribution to book/anthology

SN - 9783031553592

T3 - International Association of Geodesy Symposia

SP - 93

EP - 99

BT - X Hotine-Marussi Symposium on Mathematical Geodesy

A2 - Freymueller, Jeffrey T.

A2 - Sánchez, Laura

CY - Berlin, Heidelberg

ER -

Research@Leibniz University

Bayesian Robust Multivariate Time Series Analysis in Nonlinear Regression Models with Vector Autoregressive and t-Distributed Errors

Authors

Research Organisations

External Research Organisations

Details

Publication series

Abstract

Keywords

ASJC Scopus subject areas

Cite this

By the same author(s)

Monte Carlo variance propagation for the uncertainty modeling of a kinematic LiDAR-based multi-sensor system

Error State Kalman Filter with Implicit Measurement Equations for Position Tracking of a Multi-Sensor System with IMU and LiDAR

Automatic quality assessment of terrestrial laser scans

LUCOOP: Leibniz University Cooperative Perception and Urban Navigation Dataset

Automated damage detection for port structures using machine learning algorithms in heightfields