A spectral clustering-type algorithm for the consistent estimation of the Hurst distribution in moderately high dimensions
Abstract
Scale invariance (fractality) is a prominent feature of the large-scale behavior of many stochastic systems. In this work, we construct a provably convergent and easily interpretable algorithm for the statistical identification of the Hurst distribution (in particular, the scaling exponents) undergirding a high-dimensional fractal system. Starting from recently established properties of wavelet random matrices, the algorithm is based on modified spectral clustering. The algorithm also includes a model selection procedure so as to address the important practical problem of picking the value of the clustering precision hyperparameter. In a moderately high-dimensional regime where the dimension, the sample size and the scale go to infinity, we show that the algorithm consistently estimates the Hurst distribution. Monte Carlo simulations show that the proposed methodology is efficient for realistic sample sizes and outperforms another popular clustering method based on mixed-Gaussian modeling. We apply the algorithm in the analysis of real-world macroeconomic time series to unveil evidence of cointegration.
Related
- Consistent Spatial Clustering of Complex Objects via Bregman Geometry
- Simulation-consistent Estimation of the Marginal Likelihood for Block Models
- Strongly Consistent Estimation of the Extended ℓ1-Sum of ϕ-Mixing Coefficients from a Single Trajectory
- Consistent intercept estimation and inference for unit-root INAR(2) processes
- SENECA: Small-Sample Discrete Entropy Estimation via Self-Consistent Missing Mass
BibTeX
@article{abry2026spectral,
title = {{A spectral clustering-type algorithm for the consistent estimation of the Hurst distribution in moderately high dimensions}},
author = {Patrice Abry and Gustavo Didier and Oliver Orejola and Herwig Wendt},
journal = {Sankhya A},
year = {2026},
month = oct,
url = {https://hal.science/hal-05771961v1},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - A spectral clustering-type algorithm for the consistent estimation of the Hurst distribution in moderately high dimensions
AU - Patrice Abry
AU - Gustavo Didier
AU - Oliver Orejola
AU - Herwig Wendt
PY - 2026
DA - 2026/10/12
JO - Sankhya A
UR - https://hal.science/hal-05771961v1
AB - Scale invariance (fractality) is a prominent feature of the large-scale behavior of many stochastic systems. In this work, we construct a provably convergent and easily interpretable algorithm for the statistical identification of the Hurst distribution (in particular, the scaling exponents) undergirding a high-dimensional fractal system. Starting from recently established properties of wavelet random matrices, the algorithm is based on modified spectral clustering. The algorithm also includes a model selection procedure so as to address the important practical problem of picking the value of the clustering precision hyperparameter. In a moderately high-dimensional regime where the dimension, the sample size and the scale go to infinity, we show that the algorithm consistently estimates the Hurst distribution. Monte Carlo simulations show that the proposed methodology is efficient for realistic sample sizes and outperforms another popular clustering method based on mixed-Gaussian modeling. We apply the algorithm in the analysis of real-world macroeconomic time series to unveil evidence of cointegration.
SN - 0972-7671
SN - 0976-836X
ER -