pith. sign in

arxiv: 2604.27242 · v1 · submitted 2026-04-29 · 🧮 math.PR · math.ST· stat.TH

Statistical Inference for Homogenization Limits Driven by Wiener or Hermite Processes

Pith reviewed 2026-05-07 09:44 UTC · model grok-4.3

classification 🧮 math.PR math.STstat.TH
keywords homogenizationslow-fast systemsstatistical inferencefractional Ornstein-UhlenbeckHermite processWiener processquadratic variationsconsistency
0
0 comments X

The pith

Estimators for diffusivity and Hurst parameter remain consistent when applied to subsampled data from the original slow/fast system.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that estimators of diffusivity and Hurst parameter taken from the homogenized limit of slow/fast systems stay consistent when applied to data generated by the full original system, as long as the observations are subsampled at rates tied to the scale separation. A reader would care because real-world observations come from the multi-scale dynamics rather than the effective limit equation, so this result lets inference proceed directly from available data. The argument rests on proving L2 consistency for renormalized quadratic variations of the fast process through Wiener chaos expansions that yield an orthogonal decomposition, which carries over to both Wiener-driven and Hermite-driven limits and preserves non-central fluctuation theorems under tighter subsampling.

Core claim

We show that estimators constructed from the homogenized limit remain consistent when applied to appropriately subsampled data generated by the original slow/fast system. In the models considered the fast variable is a fractional Ornstein-Uhlenbeck process and the limit is an SDE driven by a Wiener process or a Hermite process depending on parameters. Using Wiener chaos expansions we obtain an L2-orthogonal decomposition of renormalized quadratic variations of additive functionals of the fast process, which preserves consistency and, under stricter subsampling, the non-central limit theorem for non-Gaussian fluctuations. As a consequence we also obtain consistency of an estimator for the lim

What carries the argument

Renormalized quadratic variations of additive functionals of the fractional Ornstein-Uhlenbeck fast process, equipped with an L2-orthogonal decomposition via Wiener chaos expansions.

If this is right

  • Consistency of diffusivity and Hurst estimators is preserved under the stated subsampling conditions for both Wiener and Hermite driven homogenized limits.
  • The non-central limit theorem for estimator fluctuations is preserved under stricter subsampling when fluctuations are non-Gaussian.
  • An estimator of the limiting self-similarity parameter is consistent without requiring knowledge of the limiting diffusivity.
  • The results apply directly to a class of one-dimensional fluctuation models.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same quadratic-variation machinery may extend to other fast processes whose additive functionals admit Wiener-chaos decompositions.
  • In practice one could tune the subsampling rate from data to achieve the required separation even when the scale gap is only moderate.
  • Numerical experiments on simulated slow/fast trajectories would directly test the minimal subsampling rate needed for consistency.

Load-bearing premise

The fast variable is a fractional Ornstein-Uhlenbeck process and observations satisfy subsampling rates that depend on the system parameters.

What would settle it

Generate data from a slow/fast system with fractional Ornstein-Uhlenbeck fast variable, apply the subsampling rates stated in the paper, compute the estimator for diffusivity or Hurst parameter, and observe that it fails to converge to the true value as the number of observations tends to infinity.

read the original abstract

We study the effective estimation of the diffusivity and Hurst parameter for the homogenized limit of a class of slow/fast systems. Depending on the system parameters, this limit solves a stochastic differential equation driven by either a Wiener process or a Hermite process. In the class of models we consider, the fast variable is a fractional Ornstein--Uhlenbeck process. We show that estimators constructed from the homogenized limit remain consistent when applied to appropriately subsampled data generated by the original slow/fast system. A key tool in our analysis is the consistency of renormalized quadratic variations for a family of additive functionals of the fast process. Using Wiener chaos expansions, we obtain an \(L^2\)-orthogonal decomposition of these renormalized quadratic variations. This allows us to show that, under appropriate subsampling conditions, the consistency properties of the estimators are preserved even when the data is generated by the slow/fast system rather than the homogenized limit. We also show that, under stricter subsampling conditions, a non-central limit theorem is preserved in the case where the fluctuations of the estimator around the true value are non-Gaussian. As a direct consequence of convergence in \(L^2\), we obtain consistency of an estimator for the limiting self-similarity that does not require knowledge of the limiting diffusivity. Finally, we show that our results apply to a class of one-dimensional fluctuation models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper studies statistical estimation of diffusivity and Hurst parameter in the homogenized limits of slow/fast systems with fractional Ornstein-Uhlenbeck fast variable. Depending on parameters, the limit is an SDE driven by Wiener or Hermite noise. The central result is that estimators built from the homogenized limit remain consistent when applied to data from the original system under suitable subsampling rates. The proof relies on L2-orthogonal decompositions of renormalized quadratic variations of additive functionals via Wiener chaos expansions, which transfers consistency (and, under stricter subsampling, non-central limit theorems) from the limit to the subsampled slow/fast data. A byproduct is an L2-consistent estimator of the limiting self-similarity index that does not require knowledge of the diffusivity; the results are also shown to cover a class of one-dimensional fluctuation models.

Significance. If the derivations hold, the work supplies a rigorous bridge between homogenization theory and statistical inference for multiscale systems, justifying the use of limit-based estimators on finite-scale observations. The Wiener-chaos orthogonal decomposition provides a clean, explicit control on the error terms that is a methodological strength. The parameter-independent estimator for self-similarity and the preservation of non-Gaussian fluctuations under stricter subsampling are practically useful features. The absence of ad-hoc fitted parameters or self-referential constructions in the core consistency statements adds to the result's reliability.

minor comments (3)
  1. [Abstract] Abstract: the phrase 'a family of additive functionals of the fast process' is left unspecified; adding one concrete example (e.g., the quadratic variation itself) would clarify the scope of the L2-decomposition result.
  2. [Main results] The subsampling conditions are stated to depend on system parameters; a compact table or diagram summarizing the regimes (e.g., rate thresholds for consistency vs. non-central limits) would improve readability and make the transfer arguments easier to follow.
  3. [Preliminaries] Notation for the Hermite process and its Hurst parameter should be cross-referenced to a standard definition (e.g., via the multiple Wiener-Itô integral) at first appearance to avoid ambiguity for readers outside the immediate subfield.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary, significance assessment, and recommendation of minor revision. No major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained via standard tools

full rationale

The paper establishes consistency of estimators for diffusivity and Hurst parameter by applying L2-orthogonal decompositions (via Wiener chaos expansions) to renormalized quadratic variations of the fast fractional OU process, then transferring those properties to subsampled slow/fast data under explicit parameter-dependent rates. These steps are direct consequences of the stated assumptions and expansions; no quantity is defined in terms of itself, no fitted input is relabeled as a prediction, and no load-bearing premise reduces to a self-citation or ansatz smuggled from prior work by the same author. The non-central limit preservation and self-similarity estimator follow similarly from L2 convergence without circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper relies on established properties of fractional Ornstein-Uhlenbeck processes and Wiener chaos expansions from prior literature; no new free parameters, ad-hoc axioms, or invented entities are introduced in the abstract.

axioms (2)
  • domain assumption The fast variable is a fractional Ornstein-Uhlenbeck process with standard properties
    Explicitly stated as the class of models considered.
  • standard math Wiener chaos expansions yield L2-orthogonal decompositions of renormalized quadratic variations
    Used as key technical tool for consistency proofs.

pith-pipeline@v0.9.0 · 5541 in / 1272 out tokens · 77238 ms · 2026-05-07T09:44:48.871841+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

  1. [1]

    Numerical simulation of Generalized Hermite Processes, March 2025

    Antoine Ayache, Julien Hamonier, and laurent Loosveldt. Numerical simulation of Generalized Hermite Processes, March 2025

  2. [2]

    Benedetto and Wojciech Czaja.Integration and Modern Analysis

    John J. Benedetto and Wojciech Czaja.Integration and Modern Analysis. Birkh ¨auser, Boston, 2009

  3. [3]

    Wiley-Interscience, New York Weinheim, 2nd edition edition, August 1999

    Patrick Billingsley.Convergence of Probability Measures: 316. Wiley-Interscience, New York Weinheim, 2nd edition edition, August 1999

  4. [4]

    Continuous Breuer–Major theorem: Tightness and nonstationarity.The Annals of Probability, 48(1), January 2020

    Simon Campese, Ivan Nourdin, and David Nualart. Continuous Breuer–Major theorem: Tightness and nonstationarity.The Annals of Probability, 48(1), January 2020

  5. [5]

    Fractional Ornstein-Uhlenbeck processes.Electronic Journal of Probability, 8(none):1–14, January 2003

    Patrick Cheridito, Hideyuki Kawaguchi, and Makoto Maejima. Fractional Ornstein-Uhlenbeck processes.Electronic Journal of Probability, 8(none):1–14, January 2003

  6. [6]

    Tudor, and Frederi G

    Alexandra Chronopoulou, Ciprian A. Tudor, and Frederi G. Viens. Self-similarity parameter estimation and reproduction property for non-Gaussian Hermite processes.Communications on Stochastic Analysis, 5(1), March 2011

  7. [7]

    Functional Limit Theorems for the Fractional Ornstein–Uhlenbeck Process.Journal of Theoretical Probability, 35(1):426–456, March 2022

    Johann Gehringer and Xue-Mei Li. Functional Limit Theorems for the Fractional Ornstein–Uhlenbeck Process.Journal of Theoretical Probability, 35(1):426–456, March 2022

  8. [8]

    Functional limit theorems for Volterra processes and applications to homogenization*.Nonlinearity, 35(4):1521, March 2022

    Johann Gehringer, Xue-Mei Li, and Julian Sieber. Functional limit theorems for Volterra processes and applications to homogenization*.Nonlinearity, 35(4):1521, March 2022

  9. [9]

    Springer, Berlin, Heidelberg, 1966

    Wilhelm Magnus, Fritz Oberhettinger, and Raj Pal Soni.Formulas and Theorems for the Special Functions of Mathematical Physics. Springer, Berlin, Heidelberg, 1966

  10. [10]

    Mandelbrot and John W

    Benoit B. Mandelbrot and John W. Van Ness. Fractional Brownian Motions, Fractional Noises and Applications.SIAM Review, 10(4):422–437, 1968

  11. [11]

    Vladas Pipiras and Murad S. Taqqu. Integration questions related to fractional Brownian motion. Probability Theory and Related Fields, 118(2):251–291, October 2000

  12. [12]

    Vladas Pipiras and Murad S. Taqqu. Are classes of deterministic integrands for fractional Brownian motion on an interval complete?Bernoulli, 7(6):873–897, December 2001

  13. [13]

    Vladas Pipiras and Murad S. Taqqu. Regularization and integral representations of Hermite processes. Statistics & Probability Letters, 80(23):2014–2023, December 2010

  14. [14]

    Murad S. Taqqu. Convergence of integrated processes of arbitrary Hermite rank.Zeitschrift f ¨dbend r Wahrscheinlichkeitstheorie und Verwandte Gebiete, 50(1):53–83, 1979

  15. [15]

    Variations of the fractional Brownian motion via Malliavin calculus

    Ciprian A Tudor and Frederi G Viens. Variations of the fractional Brownian motion via Malliavin calculus

  16. [16]

    Tudor and Frederi G

    Ciprian A. Tudor and Frederi G. Viens. Variations and estimators for self-similarity parameters via Malliavin calculus.The Annals of Probability, 37(6), November 2009

  17. [17]

    Veillette and Murad S

    Mark S. Veillette and Murad S. Taqqu. Properties and numerical evaluation of the Rosenblatt distribution.Bernoulli, 19(3), August 2013