Pith. sign in

REVIEW 2 major objections 4 minor 208 references

The paper develops a selfnormalized test for relevant changes in functional time series under the supremum norm, showing that a bias-corrected log-sum-exp statistic converges to a pivotal distribution depending only on the break location.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review

2026-08-01 00:48 UTC pith:3V5RLSZQ

load-bearing objection Genuinely new sup-norm selfnormalized relevant-change test with a smart bias-cancellation device, but the advertised exactness is over-broad: degenerate (cubic) extrema break the correction at the recommended smoothing rate. the 2 major comments →

arxiv 2607.27818 v1 pith:3V5RLSZQ submitted 2026-07-30 math.ST stat.MEstat.TH

Selfnormalization for relevant inference with supremum-type statistics

Patrick Bastian This is my paper
classification math.ST stat.MEstat.TH MSC 62G1062G2062M1062R10
keywords relevant hypothesis testingselfnormalizationsupremum normfunctional time serieschange point detectionlog-sum-exp smoothingbias correctionasymptotic pivotal distribution
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper tries to solve a standing problem: tests for whether a change in functional data exceeds a relevance threshold Δ, measured in the supremum norm, could not use selfnormalization because the supremum norm is not smooth and its limiting behavior depends on nuisance geometry and the long-run covariance. The proposed fix is to replace the norm by a log-sum-exp smooth approximation, work out the smoothing bias for two extreme geometries (isolated quadratic peaks and flat plateaus), and cancel that bias with a three-level linear combination of smoothers. The resulting bias-corrected statistic is asymptotically pivotal, with a limit that depends only on the unknown break fraction, so a threshold test based on it is asymptotically exact: rejection probability α0 at M=Δ, 0 inside the null, and 1 outside. If the theorem holds, this is the first selfnormalized relevant-change test in the sup-norm setting, and the smoothing-bias-correction template carries over to other supremum-type relevant hypotheses.

Core claim

The paper's central claim, Theorem 2.7, is that after estimating the break point, computing the post-estimation contrast, and applying the linear combination L_β of three log-sum-exp smoothers, the statistic sqrt(n){L_{β_n}(d̂_n) − M}/√V̂_n converges in distribution to T_{τ0}, a ratio of a standard normal to a weighted sum of squared L2 norms of two independent Brownian bridges, with weights determined by τ0. Because the limit is pivotal and free of long-run covariance, using its (1−α0)-quantile as a critical value gives an asymptotically exact test of the relevant hypotheses. The theorem rests on two explicit bias expansions: for finitely many nondegenerate interior extrema, the bias is (−l

What carries the argument

The key object is the log-sum-exp (soft maximum) functional Ψ_β(f) = (1/β) log ∫_0^1 (e^{β f(t)}+e^{−β f(t)}) dt, which tends to ||f||∞ as β→∞ and is Fréchet differentiable for finite β. Its derivative DΨ_{β,f}(h) is a weighted average of h with exponential weights, supplying a linear projection whose selfnormalized sums can be handled by a projected invariance principle. The bias corrections come from the linear combination L_β = a_1 Ψ_{c_1 β} + a_2 Ψ_{c_2 β} + a_3 Ψ_{c_3 β} with a≈(1.2509,−2.2555,2.0046) and c≈(1,1.4475,6.5232), chosen to cancel the leading bias in both geometric regimes. The asymptotic expansions are obtained via classical exponential-integral expansions, expressed throug

Load-bearing premise

The exactness claim rests on the assumption that the mean contrast's supremum is attained either at finitely many interior points with nonzero curvature or on a plateau whose near-maximal volume grows exactly like u^α; any other geometric shape of the extremal set breaks the bias-cancellation rate and the pivotal limit is not guaranteed.

What would settle it

Simulate the bias-corrected statistic at the boundary (M=Δ) for a contrast with a single interior extremum where the second derivative is zero (e.g., d(t)=1−c(t−1/2)^4) and for a contrast with the extremum at the boundary (e.g., d(t)=1−ct^2), using n=400 and β_n=n^{1/3}; if the empirical rejection rate at nominal 0.05 departs from 0.05 by more than sampling error across 10,000 replications, the geometric dichotomy is the load-bearing premise.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Applied at the boundary M=Δ, the test's rejection rate converges to the nominal level α0; for M<Δ it vanishes and for M>Δ it goes to 1.
  • Users no longer need to estimate long-run covariance or choose bandwidths; the limiting distribution depends only on the estimated break fraction τ0.
  • The bias-corrected test has non-trivial local power against radial alternatives of size n^{−1/2} in any fixed direction of the contrast.
  • The smoothing-bias expansions cover both finitely many interior quadratic extrema and positive-measure plateaus, and require no knowledge of which case holds.
  • The combination of log-sum-exp smoothing and multi-level bias cancellation applies more generally to relevant hypotheses defined by supremum-type functionals.

Where Pith is reading between the lines

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

  • The geometric-dichotomy assumption (finite interior peaks vs. plateaus with power-law near-extreme volume) is not verifiable from data; a natural extension would be a data-adaptive choice of β_n that detects which regime holds.
  • Boundary extrema are deferred in the paper; one could derive analogous one-sided exponential-integral expansions, likely changing the coefficient of (log β)/β and the recommended β_n rate.
  • The proof reduces to a projected invariance principle for linear functionals ℓ_n, suggesting the construction carries over to other nonsmooth statistics such as suprema of U-statistics or empirical processes.
  • Because the auxiliary bias-bound test is conservative and can lose power against nonconstant local alternatives, practitioners using the bias-corrected test should be aware that its exactness depends on the extremal geometry in ways the simpler bound does not.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. This paper proposes selfnormalized tests for relevant hypotheses on the sup-norm magnitude M = ||mu1 - mu2||_infinity in a change-point model for functional time series. The supremum norm is replaced by a log-sum-exp soft maximum Psi_beta, and a projected selfnormalizer based on the derivative of Psi at the estimated contrast is introduced. Explicit bias expansions are derived for two extremal geometries: finitely many interior nondegenerate extrema (Theorem 2.4) and positive-measure extremal sets with a near-extreme-volume expansion (Theorem 2.5). A three-scale linear combination of soft maxima cancels the leading bias terms without estimating geometric constants. The main result (Theorem 2.7) states that the bias-corrected statistic converges to a pivotal law depending only on the break fraction tau0, so that the boundary test has asymptotic rejection probability alpha0. Local-power analysis and simulations comparing the procedure with a bootstrap relevant-change test are also given.

Significance. Within the assumptions of Theorems 2.4 and 2.5, the paper makes a substantial contribution: it is the first selfnormalized relevant-change test in the C([0,1]) sup-norm framework, and the projected selfnormalizer cancels long-run covariance nuisance parameters in a genuinely non-smooth problem. The multiscale bias-cancellation idea is elegant and avoids direct estimation of extremal-set geometry. The paper also provides explicit bias expansions and a finite-sample correction, and the simulation study supports the qualitative claims. However, the advertised 'mild regularity conditions' cover only two geometric regimes, and the boundary exactness claim is not robust to degenerate extremal sets. A key Laplace expansion is also only sketched. These issues do not invalidate the conditional theorems, but they materially limit the broad claims made in the abstract and Section 2.2.

major comments (2)
  1. [Section 2.2 / Theorem 2.7 / Remark 2.8] The advertised exactness is not established for natural geometries outside the two regimes. The cancellation identities in (2.6)-(2.7) cover only isolated nondegenerate extrema and plateaus with F(u)=m+Ku^alpha+o(u^alpha). For a C^2 contrast such as d(t)=M-|t-1/2|^3-epsilon|t-1/2|^4, F(u)=2u^{1/3}-(2epsilon/3)u^{2/3}+o(u^{2/3}); after the constraints sum a_i/c_i=0 and sum a_i log c_i/c_i=0 kill the leading term, a residual C beta^{-4/3} sum a_i c_i^{-4/3} remains. With the recommended beta_n=n^{1/3}, sqrt(n) times this residual is O(n^{1/18}), so the statistic drifts and the boundary rejection probability is not alpha0. Smooth degenerate quartic extrema produce a non-vanishing or worse bias at the same rate. The abstract's 'mild regularity conditions' and the claim in Section 2.2 that geometry estimation is avoided are therefore too strong. Either restrict the theorem to the two geometri
  2. [Section 4.3, proof of Theorem 2.4] The Laplace expansion is not fully verified. After the change of variables y=sqrt(beta lambda_j)x, the proof states 'we omit showing that applying this expansion within the integral is valid in detail' and only lists steps (1)-(3). The expansion (2.6), including the o(beta^{-2}) term, is load-bearing: it is exactly the residual that must be o(n^{-1/2}) after bias correction in Theorem 2.7. Uniform integrability of the expansion terms and the tail estimate for |y|>r_beta need a complete argument. This is a proof gap in a central result, not merely a style issue.
minor comments (4)
  1. [Section 4.1, Theorem 4.1] The proof of the local-power theorem is abbreviated: the uniform oracle reduction under local alternatives and the uniform validity of the bias expansion along the sequence d_n are asserted rather than written out. Please expand these steps, especially the treatment of the triangular-array remainder terms.
  2. [Section 3, simulation design] The standardization factor z_rho for the AR(1) functional error process should be checked. As written, epsilon_i = z_rho(rho epsilon_{i-1}+eta_i) with eta_i Brownian bridges does not appear to give long-run variance 1/4 at t=1/2 for the stated z_rho=1/(1+rho). Please verify or correct the formula.
  3. [Remark 2.8(4)] The finite-sample correction using the constant 0.1 and the maximum pointwise standard deviation is described as conservative, but no sensitivity analysis is reported for this choice. Since the correction subtracts a nonnegligible finite-sample quantity, a short sensitivity table or a formal derivation of the domination claim would be useful.
  4. [Notation] The hat V notation is used with different bracket styles (e.g., hat V^{(\beta_n)}_n vs. hat V^{\beta_n}_n). Unify the notation for the selfnormalizer and its bias-corrected version.

Circularity Check

0 steps flagged

No significant circularity; the bias-corrected selfnormalized limit is derived from a genuine invariance principle and analytic bias expansions, with only a minor auxiliary self-citation.

full rationale

The central claim (Theorem 2.7) is not circular. The bias-correction constants are chosen by analytic constraints — 'To ensure that the zeroth order bias is unchanged while the first order bias cancels some simple algebra yields the conditions X_i a_i = 1, X_i a_i/c_i = 0, X_i a_i log c_i/c_i = 0, c_i >= 1' — which are solved once and for all from the Laplace/volume expansions in Theorems 2.4 and 2.5, and are not estimated from the data. The pivotal limit T_tau0 is obtained in Lemma 4.5, where the projected selfnormalizer cancels the nuisance scale: 'The factor sigma cancels in the ratio, so every subsequence has a further subsequence with the same limit law T_tau0.' This is a standard selfnormalization cancellation, not a fitted prediction. The only self-citation inside a proof is auxiliary: 'Arguments similar to those at the end of the proof of Theorem 3.3 in Bastian (2025) yield that for any epsilon>0 we may choose b such that tau-hat_n takes values in the interval [b,1-b] with probability 1-epsilon' — this only handles the M=0 case and is not the load-bearing step for the boundary-level claim. The paper's exactness is explicitly conditional on the geometric conditions of Theorems 2.4/2.5, and Remark 2.8 defers boundary extrema; a degenerate cubic extremum outside those conditions can make the residual bias dominate at beta=n^{1/3}. That is a correctness/coverage limitation, not a circular reduction.

Axiom & Free-Parameter Ledger

3 free parameters · 7 axioms · 0 invented entities

The central result rests on standard weak-dependence assumptions, a rate condition on the break estimator, a nondegeneracy assumption on the projected variance, and a geometric dichotomy for the contrast's extremal set. No new physical or mathematical entities are introduced. The bias-correction constants are analytic/numerical design choices, not data-fitted parameters.

free parameters (3)
  • Smoothing sequence β_n = β_n = c n^{1/3}, c∈{1,2,3} in simulations; theory allows any β_n→∞ with β_n=o(√n)
    User-chosen hyperparameter governing the bias-variance tradeoff of the softmax approximation. Asymptotic validity holds for any valid sequence, but finite-sample size and power depend on c.
  • Bias-correction constants a0, c0 = a0≈(1.2509, −2.2555, 2.0046), c0≈(1, 1.4475, 6.5232)
    Chosen in Section 2.2 by minimizing the remainder bound subject to the cancellation constraints Σ a_i=1, Σ a_i/c_i=0, Σ a_i log(c_i)/c_i=0. Not data-fitted, but the optimizer is described only numerically and not proven.
  • Finite-sample correction scale 0.1 = 0.1 in r̂²_n = 17.59β_n/2 (0.1 σ̂ log(n)/√n)²
    Ad hoc constant in Remark 2.8 to make the test more conservative in small samples; asymptotically negligible, not part of the main theorem.
axioms (7)
  • domain assumption (A1) The linearly interpolated partial-sum process U_n satisfies a weak invariance principle in C([0,1];C([0,1])) with limit B.
    Assumption 2.1(A1). Standard for mixing or physical dependence, but it is not proved for the specific error processes considered.
  • domain assumption (A2) When M>0, the estimated split satisfies |k̂n−k0|=o_p(√n).
    Assumption 2.1(A2). Cited to Hariz et al. (2007); critical for the oracle reduction that replaces estimated quantities by true break quantities.
  • domain assumption (A3) β_n→∞ and β_n=o(√n).
    Assumption 2.1(A3). Balances stochastic linearization error against deterministic smoothing bias.
  • domain assumption (A4) liminf_n σ_n² > 0, where σ_n² = E[{DΨ_{β_n,d}(B(1))}²].
    Assumption 2.1(A4). Needed so the selfnormalizer has a nondegenerate limit. The paper argues this holds under the geometric conditions but does not prove it from first principles.
  • domain assumption The contrast d has either finitely many nondegenerate interior extrema (Theorem 2.4) or a positive-measure plateau with F(u)=m+Ku^α+o(u^α) (Theorem 2.5), plus the rate conditions in Theorem 2.7.
    These geometric assumptions are the basis for the explicit bias expansions and for the cancellation that makes L_{β_n}(d)-M = o(n^{-1/2}). Boundary extrema are deferred in Remark 2.8.
  • domain assumption Uniform versions of the geometry and projected-variance conditions along local alternatives in Theorem 4.1.
    Assumed in Appendix 4.1 to extend the bias expansion and oracle reduction to the triangular array setting.
  • standard math Laplace-method asymptotic expansions are valid with controlled uniform remainder.
    Used in the proof of Theorem 2.4. The proof gives an outline of the justification (regions |y|<r_β and |y|>r_β, uniformity of Taylor expansion, exponential control), but omits the full dominated-convergence/uniform-integrability details.

pith-pipeline@v1.3.0-daily-deepseek · 21672 in / 15444 out tokens · 148018 ms · 2026-08-01T00:48:41.956232+00:00 · methodology

0 comments
read the original abstract

We develop a selfnormalized approach to inference for relevant changes in functional time series measured by the supremum norm. The main difficulty is that the supremum norm is not Hadamard differentiable, so standard projection-based selfnormalization does not apply and the limiting distribution may depend on the geometry of the extremal set and the long-run covariance structure. We address this problem by replacing the supremum norm with a smooth log-sum-exp approximation and constructing a projected selfnormalizer from its derivative. The resulting statistic has an asymptotically pivotal distribution that is free of long-run covariance nuisance parameters and depends only on the break location. We derive explicit smoothing-bias expansions for both isolated nondegenerate extrema and extremal sets of positive measure. To avoid direct estimation of geometric quantities such as the number, curvature, or measure of the extrema, we combine several smoothing levels to cancel the leading bias terms. This yields an asymptotically exact test for relevant changes under mild regularity conditions. More generally, the proposed smoothing and bias-correction principles provide a framework for combining selfnormalization with supremum-type statistics in problems involving relevant hypotheses.

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

208 extracted references · 22 canonical work pages · 2 internal anchors

  1. [1]

    Aston, J. A. D. and Kirch, C. (2012). Detecting and estimating epidemic changes in dependent functional data. Journal of Multivariate Analysis , 109:204--220

    2012

  2. [2]

    Aue, A., Rice, G., and Sonmez, O. (2018). Detecting and dating structural breaks in functional data without dimension reduction. Journal of the Royal Statistical Society (B) , 80:pp. 509--529

    2018

  3. [3]

    Bastian, P. (2025). Choosing the right norm for change point detection in functional data . Electronic Journal of Statistics , 19(2):4637 -- 4672

    2025

  4. [4]

    Bastian, P., Basu, R., and Dette, H. (2024). Multiple change point detection in functional data with applications to biomechanical fatigue data . The Annals of Applied Statistics , 18(4):3109 -- 3129

    2024

  5. [5]

    and Dette, H

    Bastian, P. and Dette, H. (2026). Gradual changes in functional time series. Journal of Time Series Analysis , 47(3):632--650

    2026

  6. [6]

    and Holzmann, H

    Berger, M. and Holzmann, H. (2026). Smooth and rough sample paths in mean derivative estimation for functional data. Scandinavian Journal of Statistics , 53(2):969--983

    2026

  7. [7]

    Berkes, I., Gabrys, R., Horv \'a th, L., and Kokoszka, P. (2009). Detecting changes in the mean of functional observations. Journal of the Royal Statistical Society (B) , 71:927--946

    2009

  8. [8]

    and Hu, Q

    Cai, L. and Hu, Q. (2024). From sparse to dense functional time series: phase transitions of detecting structural breaks and beyond. arXiv preprint, arXiv 2412.20858

    Pith/arXiv arXiv 2024

  9. [9]

    C \'a rcamo, J., Cuevas, A., and Rodr \'i guez, L.-A. (2020). Directional differentiability for supremum-type functionals: Statistical applications. Bernoulli , 26(3):2143--2175

    2020

  10. [10]

    H., Ng, W

    Chan, N. H., Ng, W. L., and Yau, C. Y. (2021). A self-normalized approach to sequential change-point detection for time series. Statistica Sinica , 31(1):491--517

    2021

  11. [11]

    and McKeague, I

    Chang, H.-w. and McKeague, I. W. (2022). Empirical likelihood-based inference for functional means with application to wearable device data. Journal of the Royal Statistical Society Series B: Statistical Methodology , 84(5):1947--1968

    2022

  12. [12]

    Chernozhukov, V., Chetverikov, D., and Kato, K. (2013). Gaussian approximations and multiplier bootstrap for maxima of sums of high-dimensional random vectors. The Annals of Statistics , 41(6):2786--2819

    2013

  13. [13]

    Dehling, H. (1983). Limit theorems for sums of weakly dependent B anach space valued random variables. Zeitschrift für Wahrscheinlichkeitstheorie und verwandte Gebiete , 63:393--432

    1983

  14. [14]

    Dette, H., Kokot, K., and Aue, A. (2020a). Functional data analysis in the Banach space of continuous functions . The Annals of Statistics , 48:1168 -- 1192

  15. [15]

    Dette, H., Kokot, K., and Volgushev, S. (2020b). Testing relevant hypotheses in functional time series via self-normalization. Journal of the Royal Statistical Society Series B: Statistical Methodology , 82(3):629--660

  16. [16]

    B., Wylie, J

    Hariz, S. B., Wylie, J. J., and Zhang, Q. (2007). Optimal rate of convergence for nonparametric change-point estimators for nonstationary sequences. Annals of Statistics , 35(4):1802--1826

    2007

  17. [17]

    and Dörnemann, N

    Kutta, T. and Dörnemann, N. (2025). Monitoring time series with short detection delay. Electronic Journal of Statistics , 19(1):2239--2275

    2025

  18. [18]

    Laplace, P. S. (1986). Memoir on the probability of the causes of events. Statistical Science , 1(3):364--378

    1986

  19. [19]

    Multiple comparison procedures for simultaneous inference in functional MANOVA

    Munko, M., Ditzhaus, M., Pauly, M., and Łukasz Smaga (2025). Multiple comparison procedures for simultaneous inference in functional MANOVA . Electronic Journal of Statistics , 19(1):2637 -- 2732

    2025

  20. [20]

    and Shang, H

    Rice, G. and Shang, H. L. (2017). A plug-in bandwidth selection procedure for long-run covariance estimation with stationary functional time series. Journal of Time Series Analysis , 38(4):591--609

    2017

  21. [21]

    Schucany, W. R. and Sommers, J. P. (1977). Improvement of kernel type density estimators. Journal of the American Statistical Association , 72(358):420--423

    1977

  22. [22]

    Schwarz, G. (1978). Estimating the dimension of a model. The Annals of Statistics , 6:461--464

    1978

  23. [23]

    Shao, X. (2010). A self-normalized approach to confidence interval construction in time series. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , 72(3):343--366

    2010

  24. [24]

    Shao, X. (2015). Self-normalization for time series: A review of recent developments. Journal of the American Statistical Association , 110(512):1797--1817

    2015

  25. [25]

    and Zhang, X

    Shao, X. and Zhang, X. (2010). Testing for change points in time series. Journal of the American Statistical Association , 105(491):1228--1240

    2010

  26. [26]

    J., Davenport, S., and Schwartzman, A

    Telschow, F. J., Davenport, S., and Schwartzman, A. (2022). Functional delta residuals and applications to simultaneous confidence bands of moment based statistics. Journal of Multivariate Analysis , 192:105085

    2022

  27. [27]

    and Kadane, J

    Tierney, L. and Kadane, J. B. (1986). Accurate approximations for posterior moments and marginal densities. Journal of the American Statistical Association , 81(393):82--86

    1986

  28. [28]

    Wang, R., Zhu, C., Volgushev, S., and Shao, X. (2022). Inference for change points in high-dimensional data via self-normalization. The Annals of Statistics , 50(2):781--806

    2022

  29. [29]

    Williams, C. K. I. and Barber, D. (1998). Bayesian classification with gaussian processes. IEEE Trans. Pattern Anal. Mach. Intell. , 20(12):1342–1351

    1998

  30. [30]

    and Shao, X

    Zhang, X. and Shao, X. (2015). Two sample inference for the second-order property of temporally dependent functional data. Bernoulli , 21(2):909--929

    2015

  31. [31]

    Xue and H

    G. Xue and H. Xu and Y. Yu , title =. Bernoulli , volume =. 2026 , doi =

    2026

  32. [32]

    Li and R

    D. Li and R. Li and H. L. Shang , title =. Annals of Statistics , volume =. 2024 , doi =

    2024

  33. [33]

    Kutta and H

    T. Kutta and H. Dette and S. Wang , title =. arXiv preprint arXiv:2511.06870 , year =

    arXiv

  34. [34]

    Bai and Q

    L. Bai and Q. Hu and W. Wu , title =. Journal of the Royal Statistical Society: Series B , pages =. 2026 , doi =

    2026

  35. [35]

    B. C. Boniece and L. Horváth and L. Trapani , title =. Journal of Econometrics , volume =. 2025 , doi =

    2025

  36. [36]

    T. G. Andersen and Y. Tan and V. Todorov and Z. Zhang , title =. Journal of Econometrics , volume =. 2025 , doi =

    2025

  37. [37]

    Journal of Time Series Analysis , volume =

    Kutta, Tim and Jach, Agnieszka and Kokoszka, Piotr , title =. Journal of Time Series Analysis , volume =. doi:https://doi.org/10.1111/jtsa.12796 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1111/jtsa.12796 , abstract =

    work page doi:10.1111/jtsa.12796

  38. [38]

    Bernoulli , number =

    Fabian Mies and Ansgar Steland , title =. Bernoulli , number =. 2023 , doi =

    2023

  39. [39]

    2025 , eprint=

    At the edge of Donsker's Theorem: Asymptotics of multiscale scan statistics , author=. 2025 , eprint=

    2025

  40. [40]

    E. S. Page , title =. Biometrika , year =

  41. [41]

    Bernoulli , number =

    Tim Kutta and Piotr Kokoszka , title =. Bernoulli , number =. 2025 , doi =

    2025

  42. [42]

    Journal of Computational and Graphical Statistics , volume =

    Hanbing Zhu and Houlin Zhou and Yehua Li and Xuejun Wang and Xinyuan Song , title =. Journal of Computational and Graphical Statistics , volume =. 2026 , publisher =. doi:10.1080/10618600.2025.2560625 , URL =

    arXiv 2026

  43. [43]

    Journal of Time Series Analysis , volume =

    Rice, Gregory and Shang, Han Lin , title =. Journal of Time Series Analysis , volume =. doi:https://doi.org/10.1111/jtsa.12229 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1111/jtsa.12229 , abstract =

    work page doi:10.1111/jtsa.12229

  44. [44]

    2025 , eprint=

    Monitoring Time Series for Relevant Changes , author=. 2025 , eprint=

    2025

  45. [45]

    The Annals of Statistics , number =

    Michael Vogt and Holger Dette , title =. The Annals of Statistics , number =. 2015 , doi =

    2015

  46. [46]

    Are deviations in a gradually varying mean relevant?

    Axel B. Are deviations in a gradually varying mean relevant?. The Annals of Statistics , number =. 2021 , nodoi =

    2021

  47. [47]

    Carlstein , title =

    E. Carlstein , title =. The Annals of Statistics , year =

  48. [48]

    Ciesielski , title =

    Z. Ciesielski , title =. Bull. Acad. Polon. Sci. , volume =. 1960 , pages =

    1960

  49. [49]

    Herrndorf , title =

    N. Herrndorf , title =. Z. Wahrscheinlichkeitstheorie verw. Gebiete , volume =. 1985 , publisher =

    1985

  50. [50]

    TCC'06 , year =

    Dwork, Cynthia and McSherry, Frank and Nissim, Kobbi and Smith, Adam , title =. TCC'06 , year =

  51. [51]

    A. W. van der Vaart and J. A. Wellner , title =. 1996 , series =

    1996

  52. [52]

    N. C. Jain and M. B. Marcus , title =. J. Funct. Anal. , volume =. 1975 , month =

    1975

  53. [53]

    2025 , eprint=

    Prokhorov Metric Convergence of the Partial Sum Process for Reconstructed Functional Data , author=. 2025 , eprint=

    2025

  54. [54]

    2026 , eprint=

    Unified theory of testing relevant hypotheses in functional time series , author=. 2026 , eprint=

    2026

  55. [55]

    Bernoulli , number =

    Weichi Wu and Zhou Zhou , title =. Bernoulli , number =. 2024 , doi =

    2024

  56. [56]

    2026 , eprint=

    Equivalence testing with data-dependent and post-hoc equivalence margins , author=. 2026 , eprint=

    2026

  57. [57]

    Frontiers in Sports and Active Living , volume =

    Quantifying and correcting for speed and stride frequency effects on running mechanics in fatiguing outdoor running , author =. Frontiers in Sports and Active Living , volume =. 2023 , month = apr, day =. doi:10.3389/fspor.2023.1085513 , pmid =

    arXiv 2023

  58. [58]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , volume =

    Khismatullina, Marina and Vogt, Michael , title =. Journal of the Royal Statistical Society Series B: Statistical Methodology , volume =. 2020 , month =. doi:10.1111/rssb.12347 , url =

    work page doi:10.1111/rssb.12347 2020

  59. [59]

    Journal of Time Series Analysis , volume =

    Bastian, Patrick and Dette, Holger , title =. Journal of Time Series Analysis , volume =. doi:https://doi.org/10.1111/jtsa.12809 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1111/jtsa.12809 , abstract =

    doi:10.1111/jtsa.12809

  60. [60]

    A note on estimating the change-point of a gradually changing stochastic process , journal =

    Alexander Aue and Josef Steinebach , keywords =. A note on estimating the change-point of a gradually changing stochastic process , journal =. 2002 , issn =. doi:https://doi.org/10.1016/S0167-7152(01)00184-5 , url =

    work page doi:10.1016/s0167-7152(01)00184-5 2002

  61. [61]

    Hušková , keywords =

    M. Hušková , keywords =. Gradual changes versus abrupt changes , journal =. 1999 , issn =. doi:https://doi.org/10.1016/S0378-3758(98)00173-6 , url =

    work page doi:10.1016/s0378-3758(98)00173-6 1999

  62. [62]

    Kania and T

    L. Kania and T. Manole and L. Wasserman and S. Balakrishnan , title =. 2025 , eprint =

    2025

  63. [63]

    2022 , eprint=

    Almost sure invariance principle of - mixing time series in Hilbert space , author=. 2022 , eprint=

    2022

  64. [64]

    Fast on-line changepoint detection using heavily-weighted CUSUM and veto-based decision rules , journal =

    Fabrizio Ghezzi and Eduardo Rossi and Lorenzo Trapani , keywords =. Fast on-line changepoint detection using heavily-weighted CUSUM and veto-based decision rules , journal =. 2025 , issn =. doi:https://doi.org/10.1016/j.jeconom.2025.106071 , url =

    arXiv 2025

  65. [65]

    2025 , eprint=

    TWIN: Two window inspection for online change point detection , author=. 2025 , eprint=

    2025

  66. [66]

    Statistica Sinica , volume =

    Dependent functional linear models with applications to monitoring structural change , author =. Statistica Sinica , volume =. 2014 , doi =

    2014

  67. [67]

    Zinn , title =

    J. Zinn , title =. Ann. Probab. , volume =. 1977 , month =

    1977

  68. [68]

    Hoffmann-Jorgensen and G

    J. Hoffmann-Jorgensen and G. Pisier , title =. Ann. Probab. , volume =. 1976 , month =

    1976

  69. [69]

    Kutta and N

    T. Kutta and N. Dörnemann , title =. Electronic Journal of Statistics , volume =. 2025 , doi =

    2025

  70. [70]

    Giraudo , title =

    D. Giraudo , title =. Braz. J. Probab. Stat. , volume =

  71. [71]

    Horváth and P

    L. Horváth and P. Kokoszka and S. Wang , title =. Ann. Statist. , volume =. 2021 , doi =

    2021

  72. [72]

    Giraudo , title =

    D. Giraudo , title =. Stochastic Process. Appl. , volume =

  73. [73]

    S. H. Weakly dependent functional data , journal =. 2010 , volume =

    2010

  74. [74]

    Messer and M

    M. Messer and M. Kirchner and J. Schiemann and J. Roeper and R. Neininger and G. Schneider , title =. The Annals of Applied Statistics , volume =

  75. [75]

    Schwarz , title =

    G. Schwarz , title =. The Annals of Statistics , volume =

  76. [76]

    Li and Q

    H. Li and Q. Guo and A. Munk , title =. Electronic Journal of Statistics , volume =

  77. [77]

    Horv \'a th and G

    L. Horv \'a th and G. Rice and L. Zhao. Change point analysis of covariance functions: a weighted cumulative sum approach. Journal of Multivariate Analysis. 2022

    2022

  78. [78]

    J. A. D. Aston and C. Kirch , title =. Journal of Multivariate Analysis , year =

  79. [79]

    Chiou and Y.-T

    J.-M. Chiou and Y.-T. Chen and T. Hsing , title =. The Annals of Applied Statistics , volume =

  80. [80]

    Rice and C

    G. Rice and C. Zhang , title =. Statistics & Probability Letters , volume =

Showing first 80 references.

This paper was first reviewed by deepseek-v4-flash on August 1, 2026.