Pith. sign in

REVIEW 3 major objections 6 minor 78 references

Matched Generators for the Karhunen--Lo\`eve Transform: A Double-Commutator Eigenvalue Theory

T0 review · 3 major / 6 minor · reviewed 2026-07-13 · grok-4.5

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

Pith's one-line read The best classical transform for a covariance is the nearest commuting generator, found by a small double-commutator eigenvalue problem whose size is independent of ambient dimension.

desk verdict Clean inverse packaging of the KLT commutant: a dim(B) double-commutator pencil plus a tridiagonal uniqueness theorem that legitimately upgrades residual minimization to exact DPSS/DCT/Hahn recovery. read the letter →

arxiv 2607.08788 v1 pith:4PLNMYAS submitted 2026-06-26 eess.SP cs.ITcs.NAmath.ITmath.NA

Mitchell A. Thornton This is my paper
classification eess.SPcs.ITcs.NAmath.ITmath.NA MSC 47A7542C0515A1894A12
keywords Karhunen–Loèvetransformdouble-commutatoreigenvalueproblemsymmetrycommutantprolatespheroidalsequencesdiscretecosineorthogonalpolynomialsgraphautomorphismsmatchedgenerators
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

The Karhunen–Loève transform is the optimal orthonormal expansion for a second-order process, yet for many real covariances it collapses onto a classical basis such as Fourier, cosine, or prolate spheroidal sequences. That collapse is forced by symmetry: when the covariance commutes with a group action, the eigenfunctions are the group’s representation functions. This paper asks the inverse question. Given only the covariance and a short list of candidate generators, which generator most nearly commutes with it? The answer is the smallest eigenvector of a double-commutator pencil whose dimension equals the number of candidates, not the ambient dimension. When an exact commuting tridiagonal generator exists, a uniqueness theorem recovers the prolate, cosine, and discrete orthogonal-polynomial bases exactly; convex combinations of generators fill in a continuum of interpolating transforms, some of them certifiably outside the sinusoidal family. Approximate symmetry has an exact coding cost equal to the multi-information among sectors, and the recovered generator remains stable under sample estimation error. The practical payoff is that a mixed covariance can be synthesized from two known generators alone, matching full-data compaction from only a handful of observations.

What carries the argument

The double-commutator pencil Mc=λGc, with M_ij=Tr(B_i^H ad_R²(B_j)) and G the Gram matrix of the generators. It converts nearest-commuting-generator search into a Hermitian generalized eigenproblem whose size is independent of ambient dimension and whose zero eigenvalue certifies exact symmetry.

What would settle it

On a covariance known to be an injective function of a given irreducible tridiagonal generator (for example the band-limiting operator or the Hahn heat kernel), solve the double-commutator pencil over the symmetric tridiagonal matrices; the second null vector must recover that generator up to affine normalization and its eigenvectors must diagonalize the covariance to machine precision. Failure of either check falsifies the uniqueness claim.

Watch

Extended reading notes

Core claim

Minimizing the normalized commutator residual δ(A,R) over a finite candidate space B is exactly the Rayleigh–Ritz problem for the double-commutator superoperator ad_R² restricted to B. Its solution is the smallest generalized eigenvector of a Hermitian positive-semidefinite pencil Mc=λGc of size dim(B). When the residual vanishes and R is an injective function of an irreducible Jacobi matrix, the tridiagonal commutant is forced to span{I,J}, so the pencil recovers J and its orthogonal-polynomial eigenbasis as the unique KLT.

Load-bearing premise

Everything recovered is only as good as the hand-chosen list of candidate generators; if the true symmetry lies outside that list, the pencil cannot find it, and exact transform recovery further requires the covariance to be an injective function of an irreducible Jacobi matrix with simple spectrum.

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

3 major / 6 minor

Summary. The paper studies the inverse problem of recovering the symmetry generator of a covariance operator R, and hence the classical or hidden transform onto which the KLT collapses. Given a finite candidate space B of generators, the generator minimizing the commutativity residual δ(A,R)=∥[R,A]∥_F/(∥R∥_F∥A∥_F) is characterized as the smallest generalized eigenvector of a double-commutator pencil Mc=λGc of size dim(B), independent of ambient dimension (Theorem 7.3). Exact symmetry is a spectral condition (λ_min=0). A tridiagonal commutant-uniqueness theorem (Theorems 12.2–12.3) upgrades residual minimization to exact recovery of the prolate spheroidal, Hahn, and DCT bases when R=f(J) for injective f and irreducible Jacobi J. Approximate symmetry yields an exact multi-information coding penalty for the blockwise group transform; permutation residuals equal eigenvalue displacement and characterize graph automorphisms; sequential deflation addresses non-Abelian symmetry under a generating-completeness hypothesis. Stability under covariance estimation and a two-paradigm synthesis application (without forming the mixed covariance) are also developed.

Significance. If the results hold as stated, the paper supplies a clean, dimension-independent variational mechanism for discovering matched transforms from second-order statistics, unifying classical commutant facts with an inverse recovery principle. The double-commutator pencil, the Jacobi uniqueness argument that turns residual minimization into transform recovery, the exact multi-information coding identity, and the permutation/automorphism formula are concrete and useful contributions. Strengths include carefully written standard derivations (self-adjointness of ad_R, Rayleigh–Ritz reduction, Davis–Kahan stability), explicit scope gates (simple spectrum, injective f, structural class of B), honest numerical-only status for Haar, open matrix-valued questions, and the documented generating-completeness failure in Appendix A. Reproducibility scripts are claimed for the numerical illustrations. The work is significant for algebraic and representation-theoretic signal processing and for inverse problems that recover structured operators from covariances.

major comments (3)
  1. [Abstract; §14, Table 2] Abstract and §14 claim that the two-paradigm synthesis reaches the full-data KLT compaction from a handful of observations on synthetic and public real-world signals. Section 14 is explicitly condensed, with the full experimental protocol deferred to a companion manuscript, and Table 2 reports only qualitative summaries (matches / improves / negligible margin) without sample sizes, variance, or held-out metrics in this text. For a signal-processing venue, either expand the experimental evidence to support the abstract claim within this manuscript, or qualify the abstract and §14 claims to match the condensed presentation (e.g., synthetic exactness plus qualitative real-data trends).
  2. [§7 Eq. (22); Theorems 12.2–12.3; Theorem 11.6, Remark 11.7, Appendix A] Exact transform recovery (Theorems 12.2–12.3) and the sequential non-Abelian procedure (Theorem 11.6) are gated by structural hypotheses on the candidate class B: B must be the right restricted class (e.g. symmetric tridiagonals of fixed bandwidth), R=f(J) for injective f on an irreducible Jacobi matrix with simple spectrum, and generating-completeness for full automorphism recovery. Appendix A shows generating-completeness can fail on C6, halting at a proper subgroup. These gates are stated, but for the central applied claim—recovering hidden transforms from data—the manuscript should more prominently discuss how a practitioner chooses or validates B, and what fails when the structural hypothesis is misspecified (beyond residual size).
  3. [Abstract; §13.1, Figure 9] Section 13.1 certifies an interpolant outside Jain’s sinusoidal family via triviality of the tridiagonal commutant, but explicitly disclaims any compaction benefit. That is fine as a structural statement; however, the abstract’s phrasing (“a continuum of transforms interpolating between and beyond the classical families”) can be read as implying practical value. A short clarification that the continuum is structural, with compaction gains demonstrated only for interior generators matched to the covariance (Figure 8), would prevent over-reading.
minor comments (6)
  1. [§1.2; References [67–71]] Dense concurrent self-citation cluster [67–71] makes it hard for a reader to see what is new here versus companion preprints. A short “relation to concurrent work” paragraph stating which theorems are original to this manuscript would help.
  2. [Figure 1] Figure 1 caption appears twice with slightly different wording; the second block looks like a duplicate typesetting artifact.
  3. [§2–§7] Notation switches between R and script R for the covariance operator; a single convention would reduce friction.
  4. [§14.2, Figure 10] In §14.2 the panel label uses N for block length while the text uses M; align notation with the rest of the paper.
  5. [Proposition 7.8; §15] Proposition 7.8’s gradient-flow realization is interesting but lightly connected to the rest; either expand the link to Section 15’s double-bracket discussion or shorten.
  6. [Figures 1–10] Several figure panels use Greek letters that render poorly in the text dump (e.g. residual δ, eigenvalues λ_k). Ensure production figures have legible math labels.

Circularity Check

1 steps flagged · score 2.0 of 10

No significant circularity in the central pencil or uniqueness theorems; mild concurrent self-citation cluster frames the broader program and condensed experiments without load-bearing the math.

  1. self citation load bearing [§1 Introduction (broader program paragraph) and §14 (companion experiments)]
    "This work is part of a broader program. The algebraic diversity paradigm for signal and data processing, from which the present results emerged, is described in [68]. The polynomial-time group-selection formula that the double-commutator pencil generalizes to the continuum is developed in [70], and the unification of the classical signal transforms under the matched-group principle is treated in [71]. The finite-dimensional blind group matching problem and its single-observation variance reduction appear in companion work [67, 69]... A fuller signal-processing treatment, with the complete expe"

    The concurrent self-citations [67–71] supply the program framing, the finite-dimensional precursor that the continuum pencil generalizes, and the fuller experiments of which §14 is a condensation. They are not used as uniqueness theorems or as the sole justification of any central proof (Theorems 7.3, 12.2–12.3 stand on classical arguments given in-paper). The burden is therefore mild and non-load-bearing for the mathematical claims; it raises the score only from 0 to 2.

full rationale

The double-commutator reduction (Lemma 7.2 + Theorem 7.3) is the standard Rayleigh–Ritz/Galerkin form of residual minimization of δ on a finite candidate space B; λ_min = 0 is an independent spectral certificate of exact commutation (Corollary 7.4, Theorem 8.1), not a definitional tautology. The tridiagonal uniqueness (Theorems 12.2–12.3) is proved from classical ingredients (simple spectrum ⇒ polynomial in R; orthogonal-polynomial bandedness ⇒ tridiagonal iff deg ≤ 1; kernel exactly span{I, J}) under the explicit hypothesis R = f(J) for injective f on an irreducible Jacobi matrix; the paper states the gates rather than smuggling them. Coding-penalty = multi-information follows from determinant/Fischer identities and Gaussian entropy (Theorems 4.9–4.11), not fitting. Exact recoveries of DPSS/Hahn/DCT are demonstrations under constructed R that satisfy the hypotheses, not predictions forced by the answer. The §14 synthesis estimates the two coefficients of a known generator span on sample covariances from few windows and evaluates held-out compaction; that is ordinary structured estimation, not fitted-input-as-prediction. The only mild circularity burden is the dense cluster of concurrent Thornton arXiv companions [67–71] that supply the algebraic-diversity framing, the finite-dimensional precursor, and the fuller experimental protocol of which §14 is a condensation; those citations are not used as uniqueness theorems or load-bearing premises for the proofs given here. Score 2 reflects that non-load-bearing self-citation cluster; the derivation chain itself is self-contained.

Assumptions & free parameters 4 free parameters · 6 assumptions · 2 invented entities

The central inverse claim rests on standard operator/linear-algebra facts plus modeling choices about the candidate generator space and structural hypotheses (simple spectrum, irreducible Jacobi, generating-completeness). No new physical entity is postulated. Free parameters are design choices (B, thresholds, mixture models), not fits that define the theorem statements.

free parameters (4)
  • Candidate generator space B (basis {B_k})
    Hand-chosen finite span (Lie algebra elements, tridiagonals, permutations, hierarchical couplings). All recovery is relative to B; wrong B cannot recover the true generator.
  • Acceptance threshold τ and iteration cap K_max
    Sequential non-Abelian procedure (§11) uses τ to accept rounded permutations; τ=0 is exact but brittle under estimation error.
  • Two-paradigm mixture weight α and diffusion time t
    Application model R(α)=α R_A+(1-α)R_B with heat kernels exp(-t G); unknown in practice and estimated indirectly via the pencil.
  • Hierarchical degeneracy-breaking δ and scale weights ρ^j
    Haar demonstration uses A=Σ ρ^j S_j + δX with chosen ρ, δ, t; numerical, not uniqueness-theorem parameters.
assumptions (6)
  • domain assumption R is Hermitian (self-adjoint); spectral results do not require positivity though covariance is the case of interest.
    Stated in introduction and §7; enables self-adjointness of ad_R under Frobenius product.
  • standard math Schur’s lemma / isotypic decomposition for group-invariant operators; Reynolds projection.
    §3–4; used to recover classical transforms and blockwise structure.
  • standard math Davis–Kahan / Weyl perturbation bounds for spectral projectors and eigenvalues.
    §4.2–4.3, §9 stability theorems.
  • domain assumption If R=f(J) with f injective on simple spectrum of irreducible Jacobi J, then tridiagonal commutant of R is span{I,J}.
    Theorem 12.2–12.3; load-bearing for exact transform recovery (prolate, Hahn, DCT).
  • ad hoc to paper Generating-completeness of B for sequential automorphism recovery.
    Theorem 11.6; not automatic—Appendix A shows a natural basis fails completeness on C6.
  • domain assumption High-resolution scalar transform coding model (rate ~ sum of marginal entropies / geometric mean of variances).
    §4.3, §6; used for multi-information coding penalty and criterion gap.
invented entities (2)
  • Commutativity residual δ(A,R) and matched-generator double-commutator pencil independent evidence
    purpose: Turns inverse symmetry search into a small Hermitian GEP independent of ambient dimension.
    Core construction of the paper; mathematically standard Rayleigh–Ritz on ad_R², packaged as the inverse KLT problem.
  • Parameterized interpolating generator family A(α)=(1-α)A1+αA2 and continuum of transforms independent evidence
    purpose: Produce transforms between and outside classical sinusoidal families with a tridiagonal-commutant certificate.
    Definition 13.1; eigenbases of convex combinations of structured self-adjoint generators.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Matched Generators for the Karhunen--Lo\`eve Transform: A Double-Commutator Eigenvalue Theory." pith.science (2026). https://pith.science/paper/4PLNMYAS

@misc{pith2026260708788,
  author       = {Pith},
  title        = {Pith review of: Matched Generators for the Karhunen--Lo\`eve Transform: A Double-Commutator Eigenvalue Theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4PLNMYAS}},
  note         = {Machine review of arXiv:2607.08788}
}
abstract

The Karhunen--Lo\`eve transform (KLT) diagonalizes the covariance of a second-order process and is optimal for mean-square truncation. Which classical transform it reduces to is governed by the symmetry commutant of the covariance: when the kernel commutes with a group action, the KLT eigenfunctions are the irreducible representation functions of that group, recovering the Fourier, cosine, Mellin, and spherical-harmonic systems. We study the inverse question. Given a covariance $R$ and a finite-dimensional space of candidate generators, the generator nearest to commuting with $R$, the minimizer of $\delta(A,R)=\|[R,A]\|_F/(\|R\|_F\|A\|_F)$, is the smallest-eigenvalue solution of a double-commutator eigenvalue problem $\mathrm{ad}_R^2(A^\ast)=\lambda A^\ast$, a Hermitian generalized eigenvalue problem of size the number of generators, independent of dimension. The framework recovers hidden transforms as well as classical ones: a variational characterization turns the existence of a commuting generator into a spectral condition, and a tridiagonal commutant-uniqueness result yields the prolate spheroidal, cosine, and discrete orthogonal-polynomial bases as exact recoveries, with matrix-valued extensions, and produces a continuum of transforms interpolating between and beyond the classical families. When symmetry is approximate, the coding penalty of the symmetry-adapted blockwise transform equals the multi-information among the sectors, an exact threshold between the fixed and data-driven transforms. We further give a graph-automorphism characterization of permutation structure, a sequential deflation for non-Abelian symmetry, and stability bounds under estimation error. As an application, the KLT of a two-paradigm covariance is synthesized from its two known generators, without forming the mixed covariance, reaching the full-data transform's compaction from few observations.

Figures

Figures reproduced from arXiv: 2607.08788 by the authors.

Figure 1
Figure 1. Continuous KL Decomposition for Mixed-Structure Covariance [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. The double-commutator eigenvalue problem for the mixed kernel. (a) Individual generator resid [PITH_FULL_IMAGE:figures/full_fig_p022_2.png] view at source ↗
Figure 4
Figure 4. Recovery of the prolate spheroidal (Slepian) transform by the double-commutator pencil ( [PITH_FULL_IMAGE:figures/full_fig_p037_4.png] view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: Exact recovery of the Hahn transform by the double-commutator pencil ( [PITH_FULL_IMAGE:figures/full_fig_p039_5.png]
Figure 6
Figure 6. Figure 6: Exact recovery of the discrete cosine transform (DCT-II) by the double-commutator pencil [PITH_FULL_IMAGE:figures/full_fig_p041_6.png]
Figure 7
Figure 7. Figure 7: Numerical recovery of the Haar wavelet basis by the double-commutator pencil ( [PITH_FULL_IMAGE:figures/full_fig_p043_7.png]
Figure 4
Figure 4. Figure 4: Interpolating Between Real Fourier and DCT via Self-Adjoint Laplacian Generators [PITH_FULL_IMAGE:figures/full_fig_p048_4.png]
Figure 8
Figure 8. Figure 8: Interpolation between the real Fourier and DCT bases via the self-adjoint Laplacian generator [PITH_FULL_IMAGE:figures/full_fig_p048_8.png]
Figure 9
Figure 9. Figure 9: An interpolated transform outside Jain’s sinusoidal family, from the cosine generator (a Jain mem [PITH_FULL_IMAGE:figures/full_fig_p049_9.png]
Figure 10
Figure 10. Figure 10: Synthesized transform versus the KLT on a Fourier and cosine mixture [PITH_FULL_IMAGE:figures/full_fig_p051_10.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

78 extracted references · 7 linked inside Pith

  1. [1]

    Natarajan, and K

    Nasir Ahmed, T. Natarajan, and K. R. Rao. Discrete cosine transform.IEEE Transactions on Comput- ers, C-23(1):90–93, 1974

    1974

  2. [2]

    Graph isomorphism in quasipolynomial time

    L ´aszl´o Babai. Graph isomorphism in quasipolynomial time. InProceedings of the 48th ACM Sympo- sium on Theory of Computing (STOC), pages 684–697, 2016

    2016

  3. [3]

    Cambridge University Press, 2nd edition, 1993

    Norman Biggs.Algebraic Graph Theory. Cambridge University Press, 2nd edition, 1993

    1993

  4. [4]

    Brockett

    Roger W. Brockett. Dynamical systems that sort lists, diagonalize matrices, and solve linear program- ming problems.Linear Algebra and its Applications, 146:79–91, 1991

    1991

  5. [5]

    Jacobi angles for simultaneous diagonalization

    Jean-Franc ¸ois Cardoso and Antoine Souloumiac. Jacobi angles for simultaneous diagonalization. SIAM Journal on Matrix Analysis and Applications, 17(1):161–164, 1996

    1996

  6. [6]

    Riley Casper, F

    W. Riley Casper, F. Alberto Gr ¨unbaum, Milen Yakimov, and Ignacio Zurri ´an. Reflective prolate- spheroidal operators and the adelic Grassmannian.Communications on Pure and Applied Mathemat- ics, 76(12):3769–3810, 2023

    2023

  7. [7]

    Riley Casper and Milen Yakimov

    W. Riley Casper and Milen Yakimov. The matrix Bochner problem.American Journal of Mathematics, 144(4):1009–1065, 2022. 54

    2022

  8. [8]

    Moody T. Chu. Linear algebra algorithms as dynamical systems.Acta Numerica, 17:1–86, 2008

    2008

Show all 78 references
  1. [9]

    Chu and Kenneth R

    Moody T. Chu and Kenneth R. Driessel. The projected gradient method for least squares matrix approximations with spectral constraints.SIAM Journal on Numerical Analysis, 27(4):1050–1060, 1990

    1990

  2. [10]

    Interscience, 1953

    Richard Courant and David Hilbert.Methods of Mathematical Physics, volume 1. Interscience, 1953

    1953

  3. [11]

    Cover and Joy A

    Thomas M. Cover and Joy A. Thomas.Elements of Information Theory. Wiley, 2nd edition, 2006

    2006

  4. [12]

    Society for Industrial and Applied Mathematics, Philadelphia, 1992

    Ingrid Daubechies.Ten Lectures on Wavelets, volume 61 ofCBMS-NSF Regional Conference Series in Applied Mathematics. Society for Industrial and Applied Mathematics, Philadelphia, 1992

    1992

  5. [13]

    Marcel F. E. de Jeu. The dunkl transform.Inventiones Mathematicae, 113(1):147–162, 1993

    1993

  6. [14]

    Institute of Mathematical Statistics, Hayward, CA, 1988

    Persi Diaconis.Group Representations in Probability and Statistics, volume 11 ofInstitute of Math- ematical Statistics Lecture Notes-Monograph Series. Institute of Mathematical Statistics, Hayward, CA, 1988

    1988

  7. [15]

    What can we do in a symmetry-constrained perspective? the importance of the total charge’s status in quantum reference frame frameworks.Quantum, 10:2126, 2026

    Guilhem Doat and Augustin Vanrietvelde. What can we do in a symmetry-constrained perspective? the importance of the total charge’s status in quantum reference frame frameworks.Quantum, 10:2126, 2026

    2026

  8. [16]

    J. J. Duistermaat and F. A. Gr ¨unbaum. Differential equations in the spectral parameter.Communica- tions in Mathematical Physics, 103(2):177–240, 1986

    1986

  9. [17]

    Charles F. Dunkl. Differential-difference operators associated to reflection groups.Transactions of the American Mathematical Society, 311(1):167–183, 1989

    1989

  10. [18]

    Dur ´an and F

    Antonio J. Dur ´an and F. Alberto Gr ¨unbaum. Orthogonal matrix polynomials satisfying second-order differential equations.International Mathematics Research Notices, (10):461–484, 2004

    2004

  11. [19]

    Suboptimality of the Karhunen–Lo `eve transform for transform coding.IEEE Transactions on Information Theory, 50(8):1605–1619, 2004

    Michelle Effros, Hanying Feng, and Kenneth Zeger. Suboptimality of the Karhunen–Lo `eve transform for transform coding.IEEE Transactions on Information Theory, 50(8):1605–1619, 2004

    2004

  12. [20]

    On a theorem of Weyl concerning eigenvalues of linear transformations

    Ky Fan. On a theorem of Weyl concerning eigenvalues of linear transformations. I.Proceedings of the National Academy of Sciences, 35(11):652–655, 1949

    1949

  13. [21]

    Resection of high frequency oscillations predicts seizure outcome in the individual patient.Scientific Reports, 7(1):13836, 2017

    Tommaso Fedele, Sergey Burnos, Ece Boran, Niklaus Krayenb ¨uhl, Peter Hilfiker, Thomas Grunwald, and Johannes Sarnthein. Resection of high frequency oscillations predicts seizure outcome in the individual patient.Scientific Reports, 7(1):13836, 2017

    2017

  14. [22]

    Interictal iEEG during slow-wave sleep with HFO markings

    Tommaso Fedele, Niklaus Krayenb ¨uhl, Peter Hilfiker, Adam Li, and Johannes Sarnthein. Interictal iEEG during slow-wave sleep with HFO markings. OpenNeuro, dataset ds003498, 2021

    2021

  15. [23]

    Bernhard N. Flury. Common principal components inkgroups.Journal of the American Statistical Association, 79(388):892–898, 1984. 55

    1984

  16. [24]

    Folland.A Course in Abstract Harmonic Analysis

    Gerald B. Folland.A Course in Abstract Harmonic Analysis. CRC Press, Boca Raton, FL, 1995

    1995

  17. [25]

    Springer, 2001

    Chris Godsil and Gordon Royle.Algebraic Graph Theory. Springer, 2001

    2001

  18. [26]

    Vivek K. Goyal. Theoretical foundations of transform coding.IEEE Signal Processing Magazine, 18(5):9–21, 2001

    2001

  19. [27]

    Robert M. Gray. Toeplitz and circulant matrices: A review.Foundations and Trends in Communica- tions and Information Theory, 2(3):155–239, 2006

    2006

  20. [28]

    Alberto Gr ¨unbaum and Luc Haine

    F. Alberto Gr ¨unbaum and Luc Haine. Bispectral Darboux transformations: An extension of the Krall polynomials.International Mathematics Research Notices, 1997(8):359–392, 1997

    1997

  21. [29]

    Alberto Gr ¨unbaum, Luigi Longhi, and Marvin Perlstadt

    F. Alberto Gr ¨unbaum, Luigi Longhi, and Marvin Perlstadt. Differential operators commuting with finite convolution integral operators: Some nonabelian examples.SIAM Journal on Applied Mathe- matics, 42(5):941–955, 1982

    1982

  22. [30]

    Alberto Gr ¨unbaum, In ´es Pacharoni, and Ignacio Zurri ´an

    F. Alberto Gr ¨unbaum, In ´es Pacharoni, and Ignacio Zurri ´an. Bispectrality and time-band limiting: Matrix-valued polynomials.International Mathematics Research Notices, 2020(13):4016–4036, 2020

    2020

  23. [31]

    Alberto Gr ¨unbaum, Luc Vinet, and Alexei Zhedanov

    F. Alberto Gr ¨unbaum, Luc Vinet, and Alexei Zhedanov. Algebraic Heun operator and band-time limiting.Communications in Mathematical Physics, 364(3):1041–1068, 2018

    2018

  24. [32]

    Alberto Gr ¨unbaum and Milen Yakimov

    F. Alberto Gr ¨unbaum and Milen Yakimov. Discrete bispectral Darboux transformations from Jacobi operators.Pacific Journal of Mathematics, 204(2):395–431, 2002

    2002

  25. [33]

    Alberto Gr ¨unbaum and Milen Yakimov

    F. Alberto Gr ¨unbaum and Milen Yakimov. The prolate spheroidal phenomenon as a consequence of bispectrality. InSuperintegrability in Classical and Quantum Systems, volume 37 ofCRM Proceedings and Lecture Notes, pages 301–312. American Mathematical Society, Providence, RI, 2004

    2004

  26. [34]

    Zur theorie der orthogonalen funktionensysteme.Mathematische Annalen, 69(3):331– 371, 1910

    Alfr ´ed Haar. Zur theorie der orthogonalen funktionensysteme.Mathematische Annalen, 69(3):331– 371, 1910

    1910

  27. [35]

    Commutative rings of difference operators and an adelic flag manifold

    Luc Haine and Plamen Iliev. Commutative rings of difference operators and an adelic flag manifold. International Mathematics Research Notices, 2000(6):281–323, 2000

    2000

  28. [36]

    Moore.Optimization and Dynamical Systems

    Uwe Helmke and John B. Moore.Optimization and Dynamical Systems. Springer, 1994

    1994

  29. [37]

    Analysis of a complex of statistical variables into principal components.Journal of Educational Psychology, 24:417–441, 498–520, 1933

    Harold Hotelling. Analysis of a complex of statistical variables into principal components.Journal of Educational Psychology, 24:417–441, 498–520, 1933

    1933

  30. [38]

    Clarkson, Misha Kilmer, Haim Avron, and Lior Horesh

    Paulina Hoyos, Shashanka Ubaru, Dongsung Huh, Vasileios Kalantzis, Kenneth L. Clarkson, Misha Kilmer, Haim Avron, and Lior Horesh. Group-algebraic tensors: Provably-optimal equivariant learning and physical symmetry discovery. arXiv:2605.20440 [cs.LG], 2026

    2026 arXiv

  31. [39]

    Huang and P

    J.-Y . Huang and P. M. Schultheiss. Block quantization of correlated Gaussian random variables.IEEE Transactions on Communication Systems, 11(3):289–296, 1963. 56

    1963

  32. [40]

    Anil K. Jain. A sinusoidal family of unitary transforms.IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-1(4):356–365, 1979

    1979

  33. [41]

    ¨Uber lineare methoden in der wahrscheinlichkeitsrechnung.Annales Academiae Scientiarum Fennicae, Ser

    Kari Karhunen. ¨Uber lineare methoden in der wahrscheinlichkeitsrechnung.Annales Academiae Scientiarum Fennicae, Ser. A.I. Math.-Phys., (37):1–79, 1947

    1947

  34. [42]

    Lesky, and Ren ´e F

    Roelof Koekoek, Peter A. Lesky, and Ren ´e F. Swarttouw.Hypergeometric Orthogonal Polynomials and Their q-Analogues. Springer Monographs in Mathematics. Springer, Berlin, 2010

    2010

  35. [43]

    D. D. Kosambi. Statistics in function space.Journal of the Indian Mathematical Society, 7:76–88, 1943

    1943

  36. [44]

    Sergei V . Kozyrev. Wavelet theory asp-adic spectral analysis.Izvestiya: Mathematics, 66(2):367–376, 2002

    2002

  37. [45]

    Harold W. Kuhn. The hungarian method for the assignment problem.Naval Research Logistics Quar- terly, 2(1–2):83–97, 1955

    1955

  38. [46]

    Landau and Henry O

    Henry J. Landau and Henry O. Pollak. Prolate spheroidal wave functions, Fourier analysis and uncertainty—II.Bell System Technical Journal, 40(1):65–84, 1961

    1961

  39. [47]

    Springer, 4th edition, 1978

    Michel Lo `eve.Probability Theory II. Springer, 4th edition, 1978

    1978

  40. [48]

    St ´ephane G. Mallat. A theory for multiresolution signal decomposition: the wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(7):674–693, 1989

    1989

  41. [49]

    Functions of positive and negative type, and their connection with the theory of integral equations.Philosophical Transactions of the Royal Society of London A, 209:415–446, 1909

    James Mercer. Functions of positive and negative type, and their connection with the theory of integral equations.Philosophical Transactions of the Royal Society of London A, 209:415–446, 1909

    1909

  42. [50]

    Moody and Roger G

    George B. Moody and Roger G. Mark. The impact of the MIT-BIH arrhythmia database.IEEE Engineering in Medicine and Biology Magazine, 20(3):45–50, 2001

    2001

  43. [51]

    The fractional order fourier transform and its application to quantum mechanics

    Victor Namias. The fractional order fourier transform and its application to quantum mechanics. Journal of the Institute of Mathematics and its Applications, 25(3):241–265, 1980

    1980

  44. [52]

    On lines and planes of closest fit to systems of points in space.Philosophical Magazine, Series 6, 2(11):559–572, 1901

    Karl Pearson. On lines and planes of closest fit to systems of points in space.Philosophical Magazine, Series 6, 2(11):559–572, 1901

    1901

  45. [53]

    Percival and Andrew T

    Donald B. Percival and Andrew T. Walden.Spectral Analysis for Physical Applications: Multitaper and Conventional Univariate Techniques. Cambridge University Press, 1993

    1993

  46. [54]

    Joint approximate diagonalization of positive definite Hermitian matrices.SIAM Journal on Matrix Analysis and Applications, 22(4):1136–1152, 2001

    Dinh Tuan Pham. Joint approximate diagonalization of positive definite Hermitian matrices.SIAM Journal on Matrix Analysis and Applications, 22(4):1136–1152, 2001

    2001

  47. [55]

    Markus P ¨uschel and Jos´e M. F. Moura. Algebraic signal processing theory: 1-D space.IEEE Trans- actions on Signal Processing, 56(8):3586–3599, 2008. 57

    2008

  48. [56]

    Markus P ¨uschel and Jos´e M. F. Moura. Algebraic signal processing theory: Cooley–Tukey type algo- rithms for DCTs and DSTs.IEEE Transactions on Signal Processing, 56(4):1502–1521, 2008

    2008

  49. [57]

    Markus P ¨uschel and Jos´e M. F. Moura. Algebraic signal processing theory: Foundation and 1-d time. IEEE Transactions on Signal Processing, 56(8):3572–3585, 2008

    2008

  50. [58]

    Nguyen, Louis Schatzki, Patrick J

    Michael Ragone, Paolo Braccia, Quynh T. Nguyen, Louis Schatzki, Patrick J. Coles, Frederic Sauvage, Martin Larocca, and M. Cerezo. Representation theory for geometric quantum machine learning. arXiv:2210.07980 [quant-ph], 2023

    2023 arXiv

  51. [59]

    Aliaksei Sandryhaila and Jos ´e M. F. Moura. Discrete signal processing on graphs.IEEE Transactions on Signal Processing, 61(7):1644–1656, 2013

    2013

  52. [60]

    Peter J. Schmid. Dynamic mode decomposition of numerical and experimental data.Journal of Fluid Mechanics, 656:5–28, 2010

    2010

  53. [61]

    Springer, New York, 1977

    Jean-Pierre Serre.Linear Representations of Finite Groups, volume 42 ofGraduate Texts in Mathe- matics. Springer, New York, 1977

    1977

  54. [62]

    Shuman, Sunil K

    David I. Shuman, Sunil K. Narang, Pascal Frossard, Antonio Ortega, and Pierre Vandergheynst. The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains.IEEE Signal Processing Magazine, 30(3):83–98, 2013

    2013

  55. [63]

    Prolate spheroidal wave functions, Fourier analysis and uncertainty—V: The discrete case.Bell System Technical Journal, 57(5):1371–1430, 1978

    David Slepian. Prolate spheroidal wave functions, Fourier analysis and uncertainty—V: The discrete case.Bell System Technical Journal, 57(5):1371–1430, 1978

    1978

  56. [64]

    David Slepian and Henry O. Pollak. Prolate spheroidal wave functions, Fourier analysis and uncertainty—I.Bell System Technical Journal, 40(1):43–63, 1961

    1961

  57. [65]

    Smith and Robert B

    Wade A. Smith and Robert B. Randall. Rolling element bearing diagnostics using the Case Western Reserve University data: A benchmark study.Mechanical Systems and Signal Processing, 64–65:100– 131, 2015

    2015

  58. [66]

    Cambridge University Press, Cambridge, 1999

    Audrey Terras.Fourier Analysis on Finite Groups and Applications, volume 43 ofLondon Mathemat- ical Society Student Texts. Cambridge University Press, Cambridge, 1999

    1999

  59. [67]

    Thornton

    Mitchell A. Thornton. Algebraic diversity: Group-theoretic spectral estimation from single observa- tions. arXiv:2604.03634 [cs.LG], 2026

    2026 arXiv

  60. [68]

    Thornton

    Mitchell A. Thornton. Algebraic diversity: Principles of a group-theoretic approach to signal process- ing. arXiv:2604.19983 [eess.SP], 2026

    2026 arXiv

  61. [69]

    Thornton

    Mitchell A. Thornton. Continuous algebraic diversity: Unifying spectral, wavelet, and time-frequency analysis via lie group actions. arXiv:2605.00848 [eess.SP], 2026

    2026 arXiv

  62. [70]

    Thornton

    Mitchell A. Thornton. Polynomial-time optimal group selection via the double-commutator eigenvalue problem. arXiv:2605.00834 [cs.LG], 2026. 58

    2026 arXiv

  63. [71]

    Thornton

    Mitchell A. Thornton. Unification of signal transform theory. arXiv:2605.11589 [eess.SP], 2026

    2026 arXiv

  64. [72]

    Tracy and Harold Widom

    Craig A. Tracy and Harold Widom. Level-spacing distributions and the Airy kernel.Communications in Mathematical Physics, 159(1):151–174, 1994

    1994

  65. [73]

    Tracy and Harold Widom

    Craig A. Tracy and Harold Widom. Level-spacing distributions and the Bessel kernel.Communications in Mathematical Physics, 161(2):289–309, 1994

    1994

  66. [74]

    Sch ¨onberger, Juan Nunez-Iglesias, Franc ¸ois Boulogne, Joshua D

    St ´efan van der Walt, Johannes L. Sch ¨onberger, Juan Nunez-Iglesias, Franc ¸ois Boulogne, Joshua D. Warner, Neil Yager, Emmanuelle Gouillart, and Tony Yu. scikit-image: image processing in Python. PeerJ, 2:e453, 2014

    2014

  67. [75]

    Cambridge University Press, 2018

    Roman Vershynin.High-Dimensional Probability: An Introduction with Applications in Data Science. Cambridge University Press, 2018

    2018

  68. [76]

    A reduction of a graph to a canonical form and an algebra arising during this reduction.Nauchno-Technicheskaya Informatsia, Ser

    Boris Weisfeiler and Andrei Leman. A reduction of a graph to a canonical form and an algebra arising during this reduction.Nauchno-Technicheskaya Informatsia, Ser. 2, (9):12–16, 1968

    1968

  69. [77]

    Williams, Ioannis G

    Matthew O. Williams, Ioannis G. Kevrekidis, and Clarence W. Rowley. A data-driven approximation of the Koopman operator: Extending dynamic mode decomposition.Journal of Nonlinear Science, 25(6):1307–1346, 2015

    2015

  70. [78]

    Gregory W. Wornell. Wavelet-based representations for the1/ffamily of fractal processes.Proceed- ings of the IEEE, 81(10):1428–1450, 1993. 59

    1993

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.