Pith. sign in

REVIEW 3 major objections 5 minor 33 references

Residual-Christoffel Sampling for Random Feature Collocation of Linear PDEs

T0 review · 3 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read Sampling collocation points from the residual Christoffel density of the operator-applied features provably stabilizes random feature PDE solves.

desk verdict A solid, useful conditioning fix for random-feature collocation, with a real but contained limitation around rank truncation. read the letter →

arxiv 2607.13382 v1 pith:57TQEE3P submitted 2026-07-15 math.NA cs.NAphysics.comp-ph

Jiale Linghu , Yangshuai Wang This is my paper · ORCID
classification math.NAcs.NAphysics.comp-ph MSC 65N3565F3562K0565D1568W20
keywords randomfeaturemethodoperator-residualsamplingChristoffelPDEcollocationresidualGramwhiteningeffectivedimensionleast-squaresconditioning
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 paper claims that the instability of random feature collocation for linear PDEs can be traced to a mismatch between the sampling measure and the geometry of the operator-applied trial space, and that this mismatch is curable by construction. It builds the interior residual Gram from the operator-applied features, retains its leading eigenspace, and uses the induced Christoffel function as a sampling density with inverse-density weights, together with a whitening map that makes the population Gram the identity on the retained subspace. Conditional on the realized trial space, the sampled whitened Gram concentrates around the identity with O(r log(r/δ)/ε²) points, giving a condition number bound independent of feature count and population anisotropy. For analytic residual kernels, the operator spectrum decays stretched-exponentially, giving a polylogarithmic ridge effective dimension. The consequence would be a principled design principle for stable strong-form collocation that transfers across geometries and operator classes.

What carries the argument

The central objects are the residual-Christoffel function K_M^L(x) = ψ(x)^T G_r^† ψ(x) and the whitening map T_r = [v_1/√λ_1, …, v_r/√λ_r], where G is the population Gram of operator-applied features and G_r^† is its rank-r truncated inverse. K measures how a point's residual row is represented in the retained eigenspace; its expectation is r, so K/r is a probability density. Sampling from K/r with weights r/(N K) makes each weighted whitened row a PSD summand of trace r/N, which is the key to the matrix-Chernoff concentration bound (N ≥ 3r ln(2r/δ)/ε²) and the resulting condition-number bound κ ≤ √((1+ε)/(1−ε)). A deterministic variant maximizes successive regularized log-determinant increm

What would settle it

Take a manufactured solution on a domain with a known singularity or a feature map deliberately band-limited so that the top-r residual subspace misses the solution's main frequency, apply the method with the heuristic rank rule, and check the relative L2 error: if the error stays large while the condition number of the whitened system stays small, the retained-subspace premise is violated. The paper's RRQR pruning to k = r gives a concrete instance where accuracy collapses despite good conditioning.

Watch

Extended reading notes

Core claim

The central discovery is that the operator-residual Gram G = E[ψψ^T] defines both the correct sampling measure and the correct solver coordinates. The residual-Christoffel function K(x) = ψ(x)^T G_r^† ψ(x) quantifies the leverage of a point in the retained residual space; sampling points proportional to K/r and weighting them by 1/K makes every sampled row an unbiased, bounded-rank-one estimator of the identity on the retained subspace, so the empirical whitened Gram concentrates about I_r. The whitening map T_r simultaneously restricts the coefficient solve to the retained residual coordinates. Thus stability is achieved not by covering the domain but by resolving the operator-induced geome

Load-bearing premise

The load-bearing assumption is that the retained rank-r subspace of the operator-residual Gram contains the trial-space directions the true solution actually needs; if the solution's residual has significant components outside that subspace (or the rank heuristic truncates them), the method produces a well-conditioned but inaccurate system, as the paper's own feature-pruning and L-shape ablation show.

Editorial extensions

If this is right

  • With O(r log(r/δ)/ε²) collocation points, the whitened interior Gram lies within ε of the identity, so the interior residual block has condition number at most √((1+ε)/(1−ε)), independent of feature count M and of the population anisotropy λ₁/λ_r.
  • For uniformly analytic residual kernels, the random-feature operator eigenvalues decay as τ_j ≤ C e^{−c j^{1/d}}, so the ridge effective dimension grows only polylogarithmically with 1/ℓ; the observed retained dimension r stays small relative to M (r/M falls from 0.44 to 0.05 as M grows from 400 to 6400).
  • Residual-space sampling plus whitening removes numerical rank deficiency and reduces iterative-solver counts from caps to tens or hundreds across Poisson, Helmholtz, variable-coefficient, convection–diffusion, and elasticity benchmarks, on square, L-shaped, and annular domains in one to three spatial dimensions.
  • The deterministic leverage-greedy rule attains the lowest median errors at the smallest sample-to-rank ratios in scalar tests, improving high-frequency Poisson error by more than an order of magnitude relative to uniform and residual-adaptive sampling.
  • Boundary-aware whitening (including a boundary Gram block) reduces the full interior-plus-boundary condition number by nearly three orders of magnitude without changing error, while the boundary weight β trades accuracy against conditioning as expected.

Reading between the lines

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

  • A direct corollary the paper leaves implicit: the same residual Gram and Christoffel weights can be reused for multiple right-hand sides sharing the operator, domain, and trial space, so the leverage-construction cost is amortized across a parameter study.
  • The finite-candidate mismatch ε_cand = ||H_P − I_r|| is the practical bottleneck between the population theorem and the discrete sampler; one testable extension is to build the candidate Gram H_P itself with the same Christoffel-weighted quadrature, which would close the gap between the concentration bound and the implemented algorithm.
  • The theory is conditional on a fixed trial space; combining residual-Christoffel sampling with trial-space enrichment (e.g., appending singularity functions) addresses the demonstrated failure mode where the retained subspace lacks required directions—the paper's own L-shape experiment shows enrichment, not point selection, fixes the error.
  • The sketched range-finder construction for large M trades retained residual subspace size against construction cost; a natural extension is a sketch size adaptive to a user-specified accuracy target, blending the dense-eigendecomposition and sketch regimes.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes operator-aware row selection and coefficient whitening for random feature collocation of linear PDEs. For a fixed realized trial space, it defines the operator-residual feature vector ψ=(Lφ_1,...,Lφ_M)^T, the reference interior Gram G=E[ψψ^T], a retained rank r, and a whitening map P_coef=T_r. The randomized method samples collocation points from the residual-Christoffel density ∥T_r^Tψ∥²/r relative to a reference measure and weights the rows inversely to that density, so that the empirical whitened interior Gram is unbiased for I_r. A deterministic greedy method selects rows by the largest regularized leverage score and is shown, locally, to maximize the log-determinant increment. Theorem 1 gives a Chernoff-type concentration bound with sample complexity N≥(3/ε²)r log(2r/δ); Corollary 1 bounds the resulting condition number. Theorems 2–4 give an operator limit, stretched-exponential spectral decay for uniformly analytic residual kernels, and a ridge-leverage concentration result. Numerical experiments span Poisson, Helmholtz, convection–diffusion, variable-coefficient, and elasticity problems; they report substantially reduced condition numbers, faster LSQR convergence, and good small-sample accuracy for the greedy rule. The paper is careful to distinguish conditioning claims from accuracy claims and to report ablations (boundary weighting, whitening, feature pruning, enrichment) that expose the limitations of rank truncation.

Significance. The residual-Christoffel construction is a principled and novel extension of Christoffel sampling to differential residuals, and the population-level concentration theory is clean and appears correct. The proofs use standard matrix concentration and are transparent. The explicit sample-complexity bound and the spectral analysis of the residual kernel are valuable contributions, and the numerical comparisons are extensive. The main theoretical guarantee, however, is established for the population construction (or for a finite candidate set that is also used for whitening), while the implemented Algorithm 1 uses a separate reference quadrature and a separately drawn candidate set; the resulting deterministic mismatch ε_cand is isolated in Remark 2 but never controlled or reported. The rank-truncation step is heuristic and the method can be well-conditioned yet inaccurate, as the paper's own Table 4 and enrichment control demonstrate. These issues are fixable within the manuscript's scope.

major comments (3)
  1. [Section 3.1, Remark 2, Theorem 1] Theorem 1 guarantees concentration about I_r only for the population sampler or when the finite candidate set is also the whitening measure. In Algorithm 1, the reference Gram G_Q is formed on Q while candidates are drawn independently from μ, so the conditional target is H_P, not I_r, and the theorem does not control ∥Ĝ⊥−I_r∥ unless ε_cand is controlled. No bound on ε_cand is given and the experiments do not report it. The abstract/conclusion phrase 'sample complexity linear in retained dimension' therefore applies to a construction that differs from the numerical algorithm. Please either prove a finite-candidate concentration bound with explicit assumptions on P and Q, or restrict the claims and add a numerical report of ε_cand.
  2. [Section 5.5, Table 4, Section 2.2] The coefficient map P_coef=T_r and the gap-based rank rule (5) restrict the coefficient solve to range(T_r). Theorems 1 and 4 say nothing about whether this retained subspace contains the directions needed for solution accuracy. The paper separates conditioning from accuracy, but the failure mode is real: Table 4 shows RRQR pruning to k=r degrades the relative L2 error from 4.1e-5 to 1.6e-2, and the L-shaped enrichment control shows that missing solution directions cannot be repaired by point selection alone. This limitation should be stated explicitly near the statement of Corollary 1 and in the conclusion, and the paper should discuss what accuracy guarantees, if any, can be made for the gap-rule choice of r.
  3. [Section 3.2, Algorithm 2] The deterministic leverage-greedy construction is a headline contribution and is reported as the most accurate method at small sample sizes, but no convergence or conditioning theorem is given for it. Proposition 1 only shows that each greedy step maximizes the one-step log-determinant increment; it does not imply a bound on the final assembled Gram or on the condition number after N steps. Given the prominence of this algorithm in the experiments, the paper should either develop a stability guarantee or explicitly state that greedy selection is heuristic and supported only by numerical evidence.
minor comments (5)
  1. [Section 5.1] The sizes and construction details of the reference quadrature Q and the candidate pool P are not reported in the general setup. Some figure captions give P for one experiment, but a reproducibility section should list Q, P, and how candidates are sampled for every benchmark.
  2. [Section 2.3, Eq. (13)] The connection between ϑ and β is stated conditionally ('If the augmented reference Gram is chosen to match... then ϑ=β'). This is confusing because β is the boundary weight in (9), while G_aug^ϑ is a reference object. Please clarify whether ϑ=β is used in the experiments or only in the remark.
  3. [Figure 8] Panel (d) is labeled only 'raw whitened'; the figure caption should state which raw system (uniform, Christoffel, or both) is compared, and the conditioning values should be readable from the panel.
  4. [Theorem 3] The analytic-kernel hypothesis is verified for sine features and for tanh under a pole-separation condition, but the experiments do not check whether the chosen a_rf and parameter support actually satisfy the condition. A short note on this would avoid overclaiming the applicability of the stretched-exponential decay result.
  5. [General] There are minor typographical issues, e.g., 'Schrödinger' with unusual spacing in the introduction, and the notation M sys/s is introduced only informally. These do not affect the substance.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the concentration bounds are genuine derivations and accuracy claims are tested externally.

full rationale

The paper's derivation chain is self-contained and does not reduce a prediction to an input. The inverse-Christoffel weights in (14) are chosen so that Lemma 1 gives E[Ĝ⊥]=I_r exactly; this is an unbiased-importance-sampling construction, not a circular inference. Theorem 1 then proves concentration (25) by a matrix Chernoff bound using the resulting uniform row bound r/N, with an explicit sample complexity (24); the conclusion is not assumed in the hypotheses. Finite-candidate mismatch is honestly separated in Remark 2 (H_P vs I_r), and the conditioning result is explicitly described in Remark 3 as a retained-whitened-block statement. Accuracy claims in Section 5 are evaluated against held-out manufactured solutions, and the ablations in Table 4 and the L-shaped enrichment control acknowledge that rank truncation and trial-space deficiency can limit accuracy independently of conditioning — a correctness/approximation limitation, not circularity. There are no load-bearing self-citations: references [2] and [8] are not by the present authors. The gap-rule risk raised by the skeptic concerns accuracy outside range(T_r), which the paper itself flags, and is not a circularity in the derivation of the Gram-concentration and conditioning results.

Assumptions & free parameters 7 free parameters · 8 assumptions · 0 invented entities

No new physical or mathematical entities postulated; the residual-Christoffel function K_LM is a derived object from the operator-residual Gram. The free parameters are experimental hyperparameters and data-dependent rank selection; the axioms are mostly standard domain assumptions plus the heuristic rank rule and finite-candidate approximation.

free parameters (7)
  • retained rank r = varies per benchmark (e.g., 24–586)
    Chosen by the gap-aware rule (5) with a 10^-10 relative eigenvalue threshold; controls the whitening subspace dimension and affects accuracy. Not optimized against true error.
  • feature bandwidth a_rf = 5,6,8,18,28 depending on benchmark
    Sets the scale of random features φ_m=tanh(w·x+b) or sine; hand-chosen per test.
  • boundary weight β = 1 default; 10^-2,10^2 varied
    Relative weight of the boundary block in the least-squares objective (9); tradeoff between accuracy and conditioning.
  • ridge parameter γ / ρ_rel = e.g., 10^-6 λ1(G) in Fig. 3
    Regularizes the Gram for the ridge score (20) and Theorem 4; hand-chosen.
  • greedy regularization λ_g = 10^-8 to 10^-1
    Regularizes G_0 in the greedy selection (21); error varies by factor ~2 over this range.
  • boundary Gram mixing α = varied 0–0.5 in ablation
    Mixes interior and boundary Grams in G_α for boundary-aware whitening.
  • candidate pool size P = e.g., P=max(30N,6000)
    Size of the finite candidate set; larger P reduces the finite-candidate mismatch.
assumptions (8)
  • domain assumption Lφ_m ∈ L²(µ) and Bφ_m ∈ L²(ν) with pointwise representatives; data f,g evaluable at collocation points.
    Section 2.1, model problem.
  • domain assumption µ and ν are normalized volume/boundary reference measures.
    Section 2.1, used to define G and boundary Gram.
  • domain assumption The reference quadrature Q represents µ exactly enough that G_Q is treated as the population Gram.
    Algorithm 1 uses G_Q; theory is population-level; mismatch not bounded.
  • domain assumption For Theorem 3, k_L(ξ,x') is analytic on a Bernstein polyellipse E_ϱ(D) for each x', with uniform bound C_k.
    Section 4.2, assumed for spectral decay.
  • domain assumption Feature parameter distribution ρ has bounded support; derivatives of σ bounded on the compact affine image; R_ζ<∞.
    Section 4.2 before Eq. (27).
  • standard math Matrix concentration inequalities (Tropp) and classical analytic-kernel eigenvalue bounds hold.
    Used in proofs of Theorems 1,4 and Theorem 3.
  • ad hoc to paper Gap-aware rank rule (5) with threshold 10^-10 λ1 chooses r before sampling.
    Section 2.2, Eq. (5); no theoretical justification for the threshold or the local gap maximization; affects the method.
  • ad hoc to paper The finite candidate set C_P drawn from µ with ϖ_q=1/P is assumed to give H_P≈I_r.
    Section 3.1, Remark 2; mismatch ε_cand^P is not bounded in the theory, yet the experiments use finite candidates.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Residual-Christoffel Sampling for Random Feature Collocation of Linear PDEs." pith.science (2026). https://pith.science/paper/57TQEE3P

@misc{pith2026260713382,
  author       = {Pith},
  title        = {Pith review of: Residual-Christoffel Sampling for Random Feature Collocation of Linear PDEs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/57TQEE3P}},
  note         = {Machine review of arXiv:2607.13382}
}
read the original abstract

Random feature collocation fixes a randomly generated trial space and determines its coefficients from a linear least-squares system. Stability then depends on whether the sampled residual equations represent the geometry induced by the differential operator. We construct an operator-aware discretization in which the operator-applied features determine both the collocation measure and a coefficient whitening map. The randomized scheme combines a residual-Christoffel density with inverse-density weights, while a deterministic scalar-row alternative maximizes successive regularized log-determinant increments. Conditional on the realized trial space, the sampled whitened interior Gram is a spectral approximation to the reference Gram on the retained residual space, with sample complexity linear in the retained dimension up to a logarithmic factor. For uniformly analytic residual kernels, the associated operator has stretched-exponentially decaying eigenvalues and ridge effective dimension that is polylogarithmic in the inverse ridge scale. Experiments on scalar and vector equations, varied geometries, and one to three spatial dimensions show that residual-space sampling and whitening produce numerically full-rank transformed systems with substantially smaller condition numbers and iteration counts. The deterministic construction attains the lowest errors at the smallest scalar sample sizes. Residual-space geometry therefore yields a principled design for stable strong-form random feature collocation.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 5 canonical work pages

  1. [1]

    Foundations of Computational Mathematics25(6), 1975–2034 (2025).https://doi.org/10.1007/ s10208-025-09738-2

    Adcock, B.: Optimal sampling for least-squares approximation. Foundations of Computational Mathematics25(6), 1975–2034 (2025).https://doi.org/10.1007/ s10208-025-09738-2

    1975

  2. [2]

    arXiv preprint arXiv:2603.18251 (2026)

    Adcock, B., Can, K., Wang, X.: Christoffel adaptive sampling for sparse random feature expansions. arXiv preprint arXiv:2603.18251 (2026)

    arXiv 2026

  3. [3]

    SIAM Journal on Scientific Computing32(3), 1217–1236 (2010).https: //doi.org/10.1137/090767911

    Avron, H., Maymounkov, P., Toledo, S.: Blendenpik: Supercharging LAPACK’s least- squares solver. SIAM Journal on Scientific Computing32(3), 1217–1236 (2010).https: //doi.org/10.1137/090767911

    doi:10.1137/090767911 2010

  4. [4]

    SIAM Journal on Numerical Analysis54(3), 1785–1812 (2016).https://doi.org/10.1137/ 15M1032375

    Bao, W., Cai, Y., Jia, X., Tang, Q.: A uniformly accurate multiscale time integrator pseudospectral method for the Dirac equation in the nonrelativistic limit regime. SIAM Journal on Numerical Analysis54(3), 1785–1812 (2016).https://doi.org/10.1137/ 15M1032375

    2016

  5. [5]

    Journal of Scientific Computing71(3), 1094–1134 (2017).https://doi.org/10.1007/s10915-016-0333-3

    Bao, W., Cai, Y., Jia, X., Tang, Q.: Numerical methods and comparison for the Dirac equation in the nonrelativistic limit regime. Journal of Scientific Computing71(3), 1094–1134 (2017).https://doi.org/10.1007/s10915-016-0333-3

    work page doi:10.1007/s10915-016-0333-3 2017

  6. [6]

    Journal of Computational Physics296, 72–89 (2015).https://doi.org/10.1016/j

    Bao, W., Jiang, S., Tang, Q., Zhang, Y.: Computing the ground state and dynamics of the nonlinear Schr¨ odinger equation with nonlocal interactions via the nonuniform FFT. Journal of Computational Physics296, 72–89 (2015).https://doi.org/10.1016/j. jcp.2015.04.045

    doi:10.1016/j 2015

  7. [7]

    Computer Methods in Ap- plied Mechanics and Engineering449, 118583 (2026).https://doi.org/10.1016/j

    van Beek, J.W., Dolean, V., Moseley, B.: Local feature filtering for scalable and well- conditioned domain-decomposed random feature methods. Computer Methods in Ap- plied Mechanics and Engineering449, 118583 (2026).https://doi.org/10.1016/j. cma.2025.118583

    arXiv 2026

  8. [8]

    Cardenas, J.M., Adcock, B., Dexter, N.: CS4ML: A general framework for active learning with arbitrary data based on Christoffel functions. In: A. Oh, T. Nau- mann, A. Globerson, K. Saenko, M. Hardt, S. Levine (eds.) Advances in Neu- ral Information Processing Systems, vol. 36, pp. 19990–20037. Curran Associates, Inc. (2023). URLhttps://proceedings.neurips...

    2023

Show all 33 references
  1. [9]

    Journal of Machine Learning 1(3), 268–298 (2022).https://doi.org/10.4208/jml.220726

    Chen, J., Chi, X., E, W., Yang, Z.: Bridging traditional and machine learning-based algorithms for solving PDEs: The random feature method. Journal of Machine Learning 1(3), 268–298 (2022).https://doi.org/10.4208/jml.220726

    2022 doi

  2. [10]

    East Asian Journal on Applied Mathematics13(3), 435–463 (2023).https://doi.org/ 10.4208/eajam.2023-065.050423

    Chen, J., E, W., Luo, Y.: The random feature method for time-dependent problems. East Asian Journal on Applied Mathematics13(3), 435–463 (2023).https://doi.org/ 10.4208/eajam.2023-065.050423

    2023 doi

  3. [11]

    The SMAI Jour- nal of Computational Mathematics3, 181–203 (2017).https://doi.org/10.5802/ smai-jcm.24

    Cohen, A., Migliorati, G.: Optimal weighted least-squares methods. The SMAI Jour- nal of Computational Mathematics3, 181–203 (2017).https://doi.org/10.5802/ smai-jcm.24

    2017

  4. [12]

    In: Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pp

    Cohen, M.B., Musco, C., Musco, C.: Input sparsity time low-rank approximation via ridge leverage score sampling. In: Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pp. 1758–1777. SIAM (2017).https: //doi.org/10.1137/1.9781611974782.115

    2017 doi

  5. [13]

    Computer Methods in Applied Me- chanics and Engineering387, 114129 (2021).https://doi.org/10.1016/j.cma.2021

    Dong, S., Li, Z.: Local extreme learning machines and domain decomposition for solving linear and nonlinear partial differential equations. Computer Methods in Applied Me- chanics and Engineering387, 114129 (2021).https://doi.org/10.1016/j.cma.2021. 114129

    2021 doi

  6. [14]

    Neurocomput- ing391, 96–118 (2020).https://doi.org/10.1016/j.neucom.2019.12.099

    Dwivedi, V., Srinivasan, B.: Physics informed extreme learning machine (PIELM)–a rapid method for the numerical solution of partial differential equations. Neurocomput- ing391, 96–118 (2020).https://doi.org/10.1016/j.neucom.2019.12.099

    2020 doi

  7. [15]

    El Alaoui, A., Mahoney, M.W.: Fast randomized kernel ridge regression with statisti- cal guarantees. In: C. Cortes, N. Lawrence, D. Lee, M. Sugiyama, R. Garnett (eds.) Advances in Neural Information Processing Systems, vol. 28, pp. 775–783. Curran Asso- ciates, Inc. (2015). UR...

    2015

  8. [16]

    Fasshauer, G.E.: Solving partial differential equations by collocation with radial ba- sis functions. In: A. Le M´ ehaut´ e, C. Rabut, L.L. Schumaker (eds.) Surface Fitting and Multiresolution Methods, pp. 131–138. Vanderbilt University Press, Nashville, TN (1997)

    1997

  9. [17]

    SIAM Journal on Scientific Computing45(4), A1971–A1994 (2023).https://doi.org/10

    Gao, Z., Yan, L., Zhou, T.: Failure-informed adaptive sampling for PINNs. SIAM Journal on Scientific Computing45(4), A1971–A1994 (2023).https://doi.org/10. 1137/22M1527763

    2023

  10. [18]

    Computer Methods in Applied Mechanics and Engineering290, 73–97 (2015).https://doi.org/10.1016/j.cma.2015.02.006

    Hampton, J., Doostan, A.: Coherence motivated sampling and convergence analysis of least-squares polynomial chaos regression. Computer Methods in Applied Mechanics and Engineering290, 73–97 (2015).https://doi.org/10.1016/j.cma.2015.02.006

    2015 doi

  11. [19]

    Applied Mathematics and Computation119(2–3), 177–186 (2001).https://doi.org/10.1016/ S0096-3003(99)00255-6

    Hon, Y.C., Schaback, R.: On unsymmetric collocation by radial basis functions. Applied Mathematics and Computation119(2–3), 177–186 (2001).https://doi.org/10.1016/ S0096-3003(99)00255-6

    2001

  12. [20]

    Kansa, E.J.: Multiquadrics—a scattered data approximation scheme with applications to computational fluid-dynamics. II. solutions to parabolic, hyperbolic and elliptic par- tial differential equations. Computers & Mathematics with Applications19(8–9), 147– 161 (1990).https://d...

    1990 doi

  13. [21]

    Bernoulli23(1), 110–133 (2017).https://doi.org/10.3150/ 15-BEJ730

    Koltchinskii, V., Lounici, K.: Concentration inequalities and moment bounds for sample covariance operators. Bernoulli23(1), 110–133 (2017).https://doi.org/10.3150/ 15-BEJ730

    2017

  14. [22]

    SIAM Journal on Mathematical Analysis15(1), 133–136 (1984).https://doi.org/10.1137/0515009

    Little, G., Reade, J.B.: Eigenvalues of analytic kernels. SIAM Journal on Mathematical Analysis15(1), 133–136 (1984).https://doi.org/10.1137/0515009

    1984 doi

  15. [23]

    SIAM Review63(1), 208–228 (2021).https://doi.org/ 10.1137/19M1274067

    Lu, L., Meng, X., Mao, Z., Karniadakis, G.E.: DeepXDE: A deep learning library for solving differential equations. SIAM Review63(1), 208–228 (2021).https://doi.org/ 10.1137/19M1274067

    2021 doi

  16. [24]

    SIAM Journal on Scientific Computing 36(2), C95–C118 (2014).https://doi.org/10.1137/120866580

    Meng, X., Saunders, M.A., Mahoney, M.W.: LSRN: A parallel iterative solver for strongly over- or underdetermined systems. SIAM Journal on Scientific Computing 36(2), C95–C118 (2014).https://doi.org/10.1137/120866580

    2014 doi

  17. [25]

    Mathematics of Computation86(306), 1913– 1947 (2017).https://doi.org/10.1090/mcom/3192

    Narayan, A., Jakeman, J.D., Zhou, T.: A Christoffel function weighted least squares algorithm for collocation approximations. Mathematics of Computation86(306), 1913– 1947 (2017).https://doi.org/10.1090/mcom/3192

    1913 doi

  18. [26]

    Ad- vances in Computational Mathematics42(4), 973–993 (2016).https://doi.org/10

    Santin, G., Schaback, R.: Approximation of eigenfunctions in kernel-based spaces. Ad- vances in Computational Mathematics42(4), 973–993 (2016).https://doi.org/10. 1007/s10444-015-9449-5

    2016

  19. [27]

    Journal of Computational and Ap- plied Mathematics445, 115830 (2024).https://doi.org/10.1016/j.cam.2024.115830

    Sun, J., Dong, S., Wang, F.: Local randomized neural networks with discontinuous galerkin methods for partial differential equations. Journal of Computational and Ap- plied Mathematics445, 115830 (2024).https://doi.org/10.1016/j.cam.2024.115830

    2024

  20. [28]

    Journal of Computational and Applied Mathematics481, 117255 (2026).https://doi.org/10.1016/j.cam.2025.117255

    Tan, L., Chen, J.: High-precision randomized preconditioned iterative methods for the random feature method. Journal of Computational and Applied Mathematics481, 117255 (2026).https://doi.org/10.1016/j.cam.2025.117255

    2026

  21. [29]

    Foundations of Computational Mathematics12(4), 389–434 (2012).https://doi.org/10.1007/ s10208-011-9099-z

    Tropp, J.A.: User-friendly tail bounds for sums of random matrices. Foundations of Computational Mathematics12(4), 389–434 (2012).https://doi.org/10.1007/ s10208-011-9099-z

    2012

  22. [30]

    Foundations and Trends in Machine Learning8(1-2), 1–230 (2015).https://doi.org/10.1561/ 2200000048

    Tropp, J.A.: An introduction to matrix concentration inequalities. Foundations and Trends in Machine Learning8(1-2), 1–230 (2015).https://doi.org/10.1561/ 2200000048

    2015

  23. [31]

    Computer Methods in Applied Mechanics and Engineering 418, 116578 (2024).https://doi.org/10.1016/j.cma.2023.116578

    Wang, Y., Dong, S.: An extreme learning machine-based method for computational PDEs in higher dimensions. Computer Methods in Applied Mechanics and Engineering 418, 116578 (2024).https://doi.org/10.1016/j.cma.2023.116578

    2024

  24. [32]

    BIT Numerical Mathematics65(1), 11 (2025).https://doi.org/10.1007/ s10543-025-01053-0

    Wenzel, T., Winkle, D., Santin, G., Haasdonk, B.: Adaptive meshfree approxima- tion for linear elliptic partial differential equations with PDE-greedy kernel meth- ods. BIT Numerical Mathematics65(1), 11 (2025).https://doi.org/10.1007/ s10543-025-01053-0

    2025

  25. [33]

    Computer Methods in Applied Mechanics and Engineering403, 115671 (2023).https://doi.org/ 10.1016/j.cma.2022.115671

    Wu, C., Zhu, M., Tan, Q., Kartha, Y., Lu, L.: A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks. Computer Methods in Applied Mechanics and Engineering403, 115671 (2023).https://doi.org/ 10.1016/j.cma.2022.115671

    2023

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.