Pith. sign in

REVIEW 2 major objections 4 minor 4 references

Function-on-function Differential Regression

T0 review · 2 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read An unknown differential operator linking paired functional curves can be estimated at the minimax-optimal rate and checked by a bootstrap test.

desk verdict Novel action-aware identification for function-on-function differential regression, but the real-data analysis is invalid because the chosen L is not invertible. read the letter →

arxiv 2506.02363 v1 pith:6YW64KNZ submitted 2025-06-03 stat.ME

Tongyu Li , Fang Yao This is my paper
classification stat.ME MSC 62G0562G0862G1062G2062F40
keywords differentialoperatorfunction-on-functionregressionfunctionaldatalearningreproducingkernelHilbertspaceminimaxoptimalitygoodness-of-fittestaction-awareidentification
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper establishes that the unknown differential operator $D$ in a function-on-function regression model $F = D(U)+\varepsilon$ can be recovered from data, not by expanding $D$ in a basis but by observing its action on functions. The key manoeuvre is to assume each predictor function $U$ solves a user-specified partial differential equation $P(u)=f$ on $\Omega$ with boundary condition $B(u)=g$ on $\Gamma$, so that $U$ is represented by its forcing and boundary data through a solution operator $A$; then $D$ is identified with the integral-type operator $T = L^{-1}\circ D\circ A$ and estimated by regularized least squares in an operator reproducing kernel Hilbert space (a Hilbert space of operators with continuous evaluation). The estimator attains the minimax rate $n^{-r/(r+1)}$ in the prediction norm $\|\cdot\|_\Sigma^2$, and a wild-bootstrap goodness-of-fit test is proved asymptotically valid and consistent. An application to the thermodynamic energy equation retains the hypothesized differential operator at the 5% level. The paper thereby turns 'find the differential law from data' into a statistically tractable estimation-and-testing problem.

What carries the argument

The load-bearing object is the action-aware identification $T=L^{-1}\circ D\circ A$, where $A$ is the solution operator of the user-specified PDE in (2) and $L$ is a fixed linear differential operator with homogeneous boundary conditions; this rewrites a differential relation as an integral operator so that reproducing kernel Hilbert space tools apply. Estimation is carried out by regularized least squares over an operator reproducing kernel Hilbert space (a Hilbert space of operators with continuous evaluation), and an operator representer theorem reduces the infinite-dimensional optimization to a finite linear system for coefficient matrices in an orthonormal basis. The spectral decay $\gamma_k\asymp k^{-r}$ of the covariance operator $\Sigma$ relative to the operator space sets the regularity parameter $r$ that controls the convergence rate and the test's power boundary.

What would settle it

Generate functional data from $F=D(U)+\varepsilon$ where the predictor curves are solutions of a different differential equation than the assumed $P(u)=f$, $B(u)=g$, then estimate $\hat D$ by the proposed method and check whether $\|\hat D-D\|_\Sigma^2$ converges to zero as $n$ grows; failure to converge confirms that the solution-class assumption is necessary, while convergence would show the identification is more robust than stated.

Watch

Extended reading notes

Core claim

The central claim is that a differential regression operator is identifiable through its action. For predictors satisfying the known relation $P(u)=f$, $B(u)=g$, the map $u\mapsto \tilde u=(P(u),B(u))$ is invertible on the solution class through the solution operator $A$, so $D$ is in one-to-one correspondence with $T=L^{-1}\circ D\circ A$; since $L\circ L^{-1}$ is the identity, the model becomes $F = L T(\tilde U)+\varepsilon$, and inference can target $T$ as an integral operator in an operator reproducing kernel Hilbert space. Minimizing the regularized empirical loss $\ell_{O;\lambda}(\tilde u,f)=\|f-LO(\tilde u)\|^2_{L^2(\Omega)}+\lambda\|O\|^2_H$ yields an estimator with a Bahadur representation, and choosing $\lambda\asymp n^{-r/(r+1)}$ gives the minimax-optimal rate $\|\hat T-T\|^2_\Sigma=O_P(n^{-r/(r+1)})$. The proposed test statistic $Q_n=n^{-1}\|S_\lambda\tilde\varepsilon\|^2$ is asymptotically normal, and the wild bootstrap provides valid critical values, making the test consistent against alternatives in the parametric family $D_\theta$.

Load-bearing premise

Predictor functions must be exact solutions of the user-specified PDE $P(u)=f$, $B(u)=g$ with $P$ and $B$ known; if the observed curves are not in this solution class, the estimated operator is not the true differential relation.

Editorial extensions

If this is right

  • With $\lambda\asymp n^{-r/(r+1)}$, the estimated differential operator achieves the fastest possible prediction-error rate; no estimator based on the same data can beat it asymptotically.
  • An expert-specified parametric family of differential operators can be checked against data: under the null the bootstrap test keeps its level, and under fixed alternatives the rejection probability tends to 1.
  • The operator representer theorem makes the method computable: the infinite-dimensional operator estimate reduces to a finite linear system in a chosen orthonormal basis.
  • The framework applies directly to physical law validation, as illustrated by the thermodynamic energy equation, where the hypothesized derivative operator is retained.
  • Because $D$ is identified through its action, the method is not limited to a fixed dictionary of differentiation features and can capture derivative information a preset library would miss.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the PDE assumption is violated but the model is still fit, $\hat D$ may act as a best-fit pseudo-differential operator: it could remain predictive without recovering the true physical relation, so practitioners should verify the solution-class assumption before interpreting recovered operators.
  • The action-aware identification suggests a natural extension to unknown $P$ and $B$: the PDE itself could be treated as a dictionary selection problem, combining equation discovery with the statistical guarantees developed here.
  • The same operator-RKHS machinery may extend to nonlinear differential relations by probing the action of $D$ on families of test functions, although the paper treats a linear $L$ throughout.
  • Applied to other sensor or reanalysis data, the goodness-of-fit test could serve as a falsification tool for candidate physical laws encoded as differential operators, such as Fourier's law or Faraday's law.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper proposes a new functional regression framework for modeling a function-on-function relation driven by an unknown differential operator D, through the model F = D(U) + ε. To make D estimable, the authors introduce an action-aware identification: assuming each predictor U solves a known PDE P(u)=f, B(u)=g with solution operator A, and choosing an auxiliary linear differential operator L, they define T = L^{-1}∘D∘A and rewrite the model as F = L T(Ũ) + ε. Estimation is performed by penalized least squares over an operator reproducing kernel Hilbert space, with an explicit representer theorem. The paper states a Bahadur representation, proves minimax optimality at rate n^{-r/(r+1)} in the Σ-norm, and establishes asymptotic normality and bootstrap validity for a goodness-of-fit test. Numerical simulations and an ERA5 thermodynamic-energy example illustrate the methodology.

Significance. If the identification step is valid, the paper introduces a genuinely new nonparametric route for learning differential operators from repeated functional data, with statistical guarantees that go beyond symbolic-regression and parametric-ODE approaches. The action-aware reparameterization is elegant, and the minimax lower bound and bootstrap results would be substantial contributions. The paper also provides precise theorem statements and reports reproducible code, which is commendable. However, the applicability of the framework depends critically on the well-posedness of the auxiliary operators A and L^{-1}; the paper does not state these conditions, and the real-data application chooses an L for which they fail. This gap must be addressed before the published claims are fully supported.

major comments (2)
  1. [Section 2.1, Eqs. (3)-(4)] The identification step T = L^{-1}∘D∘A is only meaningful if L^{-1} is a well-defined solution operator for Lu=f with u=0 on Γ, but the paper never states an invertibility or well-posedness condition for this boundary value problem. The claim after (3) that 'L∘L^{-1} turns out to be the identity map' is true only on the range of L; when L is not surjective onto L²(Ω), the transformed model (4) is not equivalent to the original model (1), and the estimator and test target a different object. Please add explicit conditions (well-posedness of L and of the PDE in (2), and U belonging to the solution class so that A∘(P,B)=I) and verify them in each application.
  2. [Section 5, Table 5] The real-data example violates the well-posedness condition just described. With L=P=d/dlog(p) on Ω=(6.3,6.9) and B=I at both endpoints, the BVP u'=f with u(6.3)=u(6.9)=0 has a solution only if ∫_{6.3}^{6.9} f = 0. Observed predictors U are not constrained to satisfy U(6.9)=U(6.3), so D1(U)=U' need not lie in the range of L; then L^{-1}(D1(U)) is undefined and the transformed model (4) is not well-defined. Consequently, the p-values in Table 5 are not covered by Theorems 3-4. This example should be replaced by a well-posed choice of L (e.g., first-order derivative with a single boundary condition), or the framework must be generalized and the theory revised.
minor comments (4)
  1. [Section 4] In the sentence after the definition of D=-∇²-ω², the eigenvalue statement Dφ_k=(kπ)²+ω² is inconsistent with the definition; with φ_k(x)=√2 cos(kπx) one obtains Dφ_k=((kπ)²-ω²)φ_k. Please correct the sign or the definition of D.
  2. [Section 4] The simulation design uses predictor eigenfunctions φ_k that are also eigenfunctions of the true differential operator D, which is a particularly favorable setting for the method. A design where the predictor basis is not aligned with D would give a more demanding test of the action-aware identification.
  3. [Example 3, Section 2.2] The kernel K∂ in Example 3 is defined with a Dirac delta δ(x-ξ), which is a distribution rather than a square-integrable kernel; please clarify in what sense this defines a valid operator RKHS, or replace it with an L² kernel.
  4. [General] The paper states that all technical proofs and auxiliary lemmas are in the Supplementary Material, but this material is not included in the arXiv preprint. Given that Theorem 2's minimax lower bound is a central claim, please make the supplement available with the revision.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the core identification T=L^{-1}∘D∘A is a bijective reparameterization, and the load-bearing theory is derived from stated assumptions rather than from a fitted constant or a self-citation chain.

full rationale

The central reparameterization in Section 2.1 defines T as L^{-1}∘D∘A and then works with F=L T(\tilde U)+ε. This is an invertible change of variables: A∘(P,B) is the identity on the solution space and L∘L^{-1} is the identity on the chosen range, so recovering T is equivalent to recovering the action of D on solutions of (2). No parameter is fitted to a subset of the data and then reused as a prediction; the estimation problem (5)-(6) is a genuine regularized least-squares problem for T, and Proposition 1 derives the finite-dimensional form rather than assuming it. Theorems 1-4 are proved from Assumptions 1-5, with rates stated as upper and lower bounds; the lower bound in Theorem 2 is a minimax argument over the same class used in the upper bound, which is standard rather than circular. The only citation to prior work of one of the authors (Tan et al. 2024) appears in a list of parametric dynamic-data-analysis methods in Section 1.1 and is not used to justify the identification, the estimator, or the test; hence it is not load-bearing. I also note a non-circular correctness concern: in Section 5, taking L=P=d/dlog(p) with two-point Dirichlet boundary conditions makes L^{-1} well-defined only on functions with zero integral, so the displayed real-data p-values should be interpreted with caution. This is a well-posedness issue, not a circularity, and does not affect the self-containedness of the methodological derivation.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The central claim rests on a chain of modeling assumptions: the predictor functions must satisfy a known PDE, the auxiliary operator L must be invertible, and a set of standard statistical conditions (Assumptions 1-5) must hold. No new physical entities are introduced. The only tuned quantities are lambda, the kernel bandwidth h, and the basis truncation p, which are standard computational choices.

free parameters (4)
  • lambda
    Regularization tuning parameter in (5)-(6), selected by generalized cross-validation in Section 4.
  • h (Gaussian kernel bandwidth) = 0.01
    Hand-chosen bandwidth in the kernel K1 for all simulations and the real data example, Section 4.
  • p (basis truncation) = 10 for simulations
    Number of orthonormal basis functions in the finite-dimensional representation (7); unspecified for the real data example.
  • theta (parametric null parameter) = estimated by least squares
    Parameter in D_theta = -theta*nabla^2 used for the goodness-of-fit test and as a comparison estimator, Section 4.
assumptions (7)
  • domain assumption The predictor functions U_i are solutions of a known PDE P(u)=f, B(u)=g with predetermined operators P and B, so A∘(P,B) is the identity.
    Used in Section 2.1 (equation (2)) to transform D into T=L^{-1}∘D∘A; if this fails, D(u) does not equal L T(\tilde u).
  • domain assumption The auxiliary linear differential operator L admits a solution operator L^{-1} for the boundary value problem Lu=f, u=0 on Gamma.
    Required for the transformation (3) and for L∘L^{-1}=identity; not verified for first-order L with two-point Dirichlet conditions in the real example.
  • domain assumption Assumption 1: the error epsilon has finite Orlicz psi1 norm and its covariance is bounded above and below by constants.
    Stated in Section 3, used for probability bounds in Theorems 1-4.
  • domain assumption Assumption 2: the covariance operator Sigma is compact and the random variable ||L O(\tilde U)|| has sub-Gaussian tails.
    Used for simultaneous diagonalization in Proposition 2 and for the Newton-Raphson step.
  • domain assumption Assumption 3: the eigenvalues gamma_k of Sigma with respect to H decay as k^{-r}, r>1.
    Determines the minimax rate n^{-r/(r+1)}; a smoothness assumption standard in RKHS regression.
  • domain assumption Assumptions 4-5: the parametric family D_theta is Lipschitz in theta and the null estimator theta-hat is n^{-1/2} consistent.
    Used for the asymptotics of the goodness-of-fit test in Theorems 3-4.
  • standard math H is an operator reproducing kernel Hilbert space with bounded evaluation, and the evaluation functional O -> L O(\tilde U) is continuous.
    Needed for the representer theorem (Proposition 1) and for the loss (6) to be well-defined; not explicitly stated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Function-on-function Differential Regression." pith.science (2026). https://pith.science/paper/6YW64KNZ

@misc{pith2026250602363,
  author       = {Pith},
  title        = {Pith review of: Function-on-function Differential Regression},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6YW64KNZ}},
  note         = {Machine review of arXiv:2506.02363}
}
read the original abstract

Function-on-function regression has been a topic of substantial interest due to its broad applicability, where the relation between functional predictor and response is concerned. In this article, we propose a new framework for modeling the regression mapping that extends beyond integral type, motivated by the prevalence of physical phenomena governed by differential relations, which is referred to as function-on-function differential regression. However, a key challenge lies in representing the differential regression operator, unlike functions that can be expressed by expansions. As a main contribution, we introduce a new notion of model identification involving differential operators, defined through their action on functions. Based on this action-aware identification, we are able to develop a regularization method for estimation using operator reproducing kernel Hilbert spaces. Then a goodness-of-fit test is constructed, which facilitates model checking for differential regression relations. We establish a Bahadur representation for the regression estimator with various theoretical implications, such as the minimax optimality of the proposed estimator, and the validity and consistency of the proposed test. To illustrate the effectiveness of our method, we conduct simulation studies and an application to a real data example on the thermodynamic energy equation.

Figures

Figures reproduced from arXiv: 2506.02363 by the authors.

Figure 1
Figure 1. Monte Carlo densities of the test statistic (solid line) and its bootstrap for several [PITH_FULL_IMAGE:figures/full_fig_p026_1.png] view at source ↗
Figure 2
Figure 2. Temperatures versus logarithm of pressure in an ERA5 dataset. Each solid line [PITH_FULL_IMAGE:figures/full_fig_p028_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

4 extracted references · 3 canonical work pages

  1. [1]

    Functional Data Analysis on Wearable Sensor Data: A Systematic Review

    Acar-Denizli, N. & Delicado, P. (2024), ‘Functional data analysis on wearable sensor data: A systematic review’,arXiv preprint arXiv:2410.11562. Arbogast, T. & Bona, J. L. (2025),Functional analysis for the applied mathematician, CRC Press. Bongard, J. & Lipson, H. (2007), ‘Automated reverse engineering of nonlinear dynamical systems’,Proceedings of the N...

    work page Pith review arXiv 2024

  2. [130]

    & Zech, J

    Reinhardt, N., Wang, S. & Zech, J. (2024), ‘Statistical learning theory for neural operators’, arXiv preprint arXiv:2412.17582. R¨ othlisberger, M. & Papritz, L. (2023), ‘Quantifying the physical processes leading to atmospheric hot extremes at a global scale’,Nature Geoscience16(3), 210–216. Rudy, S. H., Brunton, S. L., Proctor, J. L. & Kutz, J. N. (2017...

    arXiv 2024

  3. [587]

    Morro, A., Giorgi, C. et al. (2023),Mathematical modelling of continuum physics, Springer. Raissi, M. (2018), ‘Deep hidden physics models: Deep learning of nonlinear partial differ- ential equations’,Journal of Machine Learning Research19(25), 1–24. Ramsay, J. O. & Hooker, G. (2017),Dynamic data analysis: modeling data with differential equations, Springe...

    work page 2023

  4. [2351]

    & Carroll, R

    Xun, X., Cao, J., Mallick, B., Maity, A. & Carroll, R. J. (2013), ‘Parameter estimation of partial differential equation models’,Journal of the American Statistical Association 108(503), 1009–1020. Yang, S., Wong, S. W. & Kou, S. (2021), ‘Inference of dynamic systems from noisy and sparse data via manifold-constrained gaussian processes’,Proceedings of th...

    work page 2013

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.