Identifying high-dimensional self-similarity based on spectral clustering applied to large wavelet random matrices - Computational Imaging and Vision
Conference Papers Year : 2024

Identifying high-dimensional self-similarity based on spectral clustering applied to large wavelet random matrices

Oliver Orejola
  • Function : Author
  • PersonId : 1151433
Gustavo Didier
  • Function : Author
  • PersonId : 1113836
Herwig Wendt
Patrice Abry

Abstract

In the modern world, systems are routinely monitored by multiple sensors. This generates "Big Data" in the form of large collections of time series. On the other hand, scaleinvariant (fractal) systems are of great interest in signal processing since they naturally emerge in several fields of application. In this work, we put forward an algorithm for the statistical identification of high-dimensional self-similarity. In the threefold limit as dimension, sample size and scale go to infinity, the method builds upon spectral clustering applied to large wavelet random matrices to consistently estimate the Hurst modes (exponents) and, hence, characterize high-dimensional self-similarity. Monte Carlo simulations show that the proposed methodology is efficient for realistic sample sizes.
Fichier principal
Vignette du fichier
MVSSEUSIPCO2024.pdf (768.26 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04699731 , version 1 (17-09-2024)

Identifiers

  • HAL Id : hal-04699731 , version 1

Cite

Oliver Orejola, Gustavo Didier, Herwig Wendt, Patrice Abry. Identifying high-dimensional self-similarity based on spectral clustering applied to large wavelet random matrices. European Signal Processing Conference (EUSIPCO), Aug 2024, Lyon, France. ⟨hal-04699731⟩
42 View
15 Download

Share

More