Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Multiscale approximate eigenvectors for multivariate self-similarity estimation

Wendt, Herwig ; Didier, Gustavo ; Carlsson, Marcus LU ; Troedsson, Erik LU and Abry, Patrice (2025) 33rd European Signal Processing Conference, EUSIPCO 2025 In European Signal Processing Conference p.2617-2621
Abstract

Many real-world systems generate multivariate time series data, often exhibiting self-similarity and scale-invariance across different modalities. The estimation of Hurst exponents in such settings is crucial for analyzing long-range dependencies. Yet, traditional eigenanalysis-based methods suffer from scale-dependent distortions, particularly in the presence of scaling amplitude discrepancies. In this work, we propose a novel multiscale eigenanalysis approach that leverages joint diagonalization of wavelet random matrices to improve estimation accuracy. By approximating a common eigenvector basis across multiple scales, our method mitigates the limitations of scale-wise eigenvalue regressions and provides robust estimation of... (More)

Many real-world systems generate multivariate time series data, often exhibiting self-similarity and scale-invariance across different modalities. The estimation of Hurst exponents in such settings is crucial for analyzing long-range dependencies. Yet, traditional eigenanalysis-based methods suffer from scale-dependent distortions, particularly in the presence of scaling amplitude discrepancies. In this work, we propose a novel multiscale eigenanalysis approach that leverages joint diagonalization of wavelet random matrices to improve estimation accuracy. By approximating a common eigenvector basis across multiple scales, our method mitigates the limitations of scale-wise eigenvalue regressions and provides robust estimation of multivariate self-similarity parameters. We demonstrate the effectiveness of our approach through extensive Monte Carlo simulations, showcasing improved performance over traditional methods in both orthogonal and non-orthogonal mixing scenarios. These findings establish joint eigenvector-based wavelet analysis as a powerful tool for multivariate self-similarity estimation.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
Hurst exponent, joint diagonalization, multivariate self-similarity, random matrices, wavelets
host publication
2025 33rd European Signal Processing Conference, EUSIPCO 2025 - Proceedings
series title
European Signal Processing Conference
pages
5 pages
publisher
European Signal Processing Conference, EUSIPCO
conference name
33rd European Signal Processing Conference, EUSIPCO 2025
conference location
Palermo, Italy
conference dates
2025-09-08 - 2025-09-12
external identifiers
  • scopus:105029865734
ISSN
2219-5491
ISBN
9789464593624
DOI
10.23919/EUSIPCO63237.2025.11226404
language
English
LU publication?
yes
additional info
Publisher Copyright: © 2025 European Signal Processing Conference, EUSIPCO. All rights reserved.
id
84efbe67-a470-4f51-84ac-fbc4d095559d
date added to LUP
2026-03-10 14:41:10
date last changed
2026-03-10 14:41:55
@inproceedings{84efbe67-a470-4f51-84ac-fbc4d095559d,
  abstract     = {{<p>Many real-world systems generate multivariate time series data, often exhibiting self-similarity and scale-invariance across different modalities. The estimation of Hurst exponents in such settings is crucial for analyzing long-range dependencies. Yet, traditional eigenanalysis-based methods suffer from scale-dependent distortions, particularly in the presence of scaling amplitude discrepancies. In this work, we propose a novel multiscale eigenanalysis approach that leverages joint diagonalization of wavelet random matrices to improve estimation accuracy. By approximating a common eigenvector basis across multiple scales, our method mitigates the limitations of scale-wise eigenvalue regressions and provides robust estimation of multivariate self-similarity parameters. We demonstrate the effectiveness of our approach through extensive Monte Carlo simulations, showcasing improved performance over traditional methods in both orthogonal and non-orthogonal mixing scenarios. These findings establish joint eigenvector-based wavelet analysis as a powerful tool for multivariate self-similarity estimation.</p>}},
  author       = {{Wendt, Herwig and Didier, Gustavo and Carlsson, Marcus and Troedsson, Erik and Abry, Patrice}},
  booktitle    = {{2025 33rd European Signal Processing Conference, EUSIPCO 2025 - Proceedings}},
  isbn         = {{9789464593624}},
  issn         = {{2219-5491}},
  keywords     = {{Hurst exponent; joint diagonalization; multivariate self-similarity; random matrices; wavelets}},
  language     = {{eng}},
  pages        = {{2617--2621}},
  publisher    = {{European Signal Processing Conference, EUSIPCO}},
  series       = {{European Signal Processing Conference}},
  title        = {{Multiscale approximate eigenvectors for multivariate self-similarity estimation}},
  url          = {{http://dx.doi.org/10.23919/EUSIPCO63237.2025.11226404}},
  doi          = {{10.23919/EUSIPCO63237.2025.11226404}},
  year         = {{2025}},
}