RECOVERY FUNCTION FOR SOBOLEV SPACES WITH DOMINATING MIXED DERIVATIVE

Linear approximation is the approximation of a function from a certain class by elements of a fixed finitedimensional subspace of that same class. For instance, for one-dimensional periodic functions, such elements are trigonometric polynomials. In the multidimensional case of interest, for functions periodic in each variable, this subspace is the set of trigonometric polynomials with a spectrum from a step hyperbolic cross. However, the question of selecting coefficients for these polynomials arises. This paper presents an apparatus for recovering functions from Sobolev spaces with a dominating mixed derivative based on given points, and establishes error estimates for the recovery. The method is based on constructing a recovery function in the form of a polynomial with a spectrum from a step hyperbolic cross, where the coefficients are calculated using the given points. The approximation error is of the order of the orthowidth, which is an optimal result for such polynomials. The proposed method is exact for polynomials with a spectrum from a step hyperbolic cross. Additionally, a functional that recovers the Fourier coefficients for functions from the indicated spaces is derived. Due to the explicit expression of the recovery function, the obtained formula can be used to solve applied problems.

1. Temlyakov, V.N. Approximations of functions with bounded mixed derivative, Proc. Steklov Inst. Math., 178, 3–113 (1986).

2. Temlyakov, V.N. Widths of some classes of functions of several variables, Soviet Math Dokl., 267(2), 314–317 (1982).

3. Temlyakov, V.N. Approximation of periodic functions of several variables by trigonometric polynomials, and widths of some classes of functions, Math. USSR-Izv., 49(5), 986–1030 (1985). https://doi.org/10.1070/IM1986v027n02ABEH001179.

4. Temlyakov, V.N. Estimates for the asymptotic characteristics of classes of functions with bounded mixed derivative or difference, Proc. Steklov Inst. Math., 189, 138–168 (1989).

5. Temlyakov, V.N. On Estimates of ε-Entropy and Widths of Classes of Functions with Bounded Mixed Derivative or Difference, Soviet Math. Dokl., 301, 288–291 (1988).

6. Galeev, A.М., Approximation by Fourier sums of classes of functions with several bounded derivatives, Math. Notes, 23(2), 109–117 (1978), ). https://doi.org/10.1007/BF01153149. (in Russian).

7. Nikolskaya, N.S. Approximation of differentiable functions of several variables by Fourier sums in the Lp metric. Siberian Math. J.,15 (2), 282–295 (1974). https://doi.org/10.1007/BF00968291. (in Russian).

8. Byrenheid, G. and Ullrich, T. Optimal sampling recovery of mixed order Sobolev embeddings via discrete Littlewood–Paley type characterizations, Anal Math., 43, 133–191 (2017). https://doi.org/10.1007/s10476-017-0303-5

9. Dinh Dung and Ullrich T., Lower bounds for the integration error for multivariate functions with mixed smoothness and optimal Fibonacci cubature for functions on the square, Math. Nachr., 288, 743–762 (2015). https://doi.org/10.1002/mana.201400048.

10. Krieg, D. et al. Sampling recovery in $ L_2 $ and other norms, arXiv preprint arXiv:2305.07539, 2023.

11. Kolomoitsev, Y., Lomako, T. and Tikhonov, S. Sparse grid approximation in weighted Wiener spaces, J. Fourier Anal. Appl., 29 (2), 1–19 (2023). https://doi.org/10.1007/s00041-023-09994-2.

12. Byrenheid, G., Sparse Representation of Multivariate Functions Based on Discrete Point Evaluations, Universitäts-und Landesbibliothek Bonn, 2019.

13. Kolomoitsev, Y. Approximation by quasi-interpolation operators and Smolyak's algorithm, J. Complexity, 69, 101601 (2022). https://doi.org/10.1016/j.jco.2021.101601.

14. Dinh Dung B-spline quasi-interpolant representations and sampling recovery of functions with mixed smoothness, J. Complexity, 27, 541–467 (2011). https://doi.org/10.1016/j.jco.2011.02.004.

15. Triebel, H. Bases in Function Spaces, Sampling, Discrepancy, Numerical Integration, European Math. Soc. Publishing House, Zürich, 2010.

16. Krieg, D. and Ullrich, M. Function values are enough for L 2-approximation Found. Comput. Math., 21(4), 1141–1151 (2021). https://doi.org/10.1007/s10208-020-09481-w.

17. Nursultanov, E.D., Tleukhanova, N.T. On the reconstruction of multiplicative transformations of functions from anisotropic spaces, Siberian Math. J., 55(3), 482–497 (2014). https://doi.org/10.1134/S0037446614030100 (in Russian).

18. Dũng, D., Temlyakov, V., Ullrich, T. Hyperbolic Cross Approximation. Springer, 2018.

19. Nursultanov, E.D. On the multipliers of Fourier series with respect to the trigonometric system, Math. Notes, 63(2), 205–214 (1998). https://doi.org/10.4213/mzm1270.

20. Bassarov, S., Nursultanov, E. New Cubature Formulas for Sobolev Spaces with Dominant Mixed Derivative. J Math Sci., 291, 4–19 (2025). https://doi.org/10.1007/s10958-025-07774-5

21. Dick, J., Kritzer, P., Pillichshammer, F. Lattice Rules. Springer. Cham. 2022. https://doi.org/10.1007/978-3-031-09951-9.

22. Plonka, G. et al. Numerical Fourier Analysis. Springer International Publishing. New York. 2018. https://doi.org/10.1007/978-3-031-35005-4.

23. Bartel, F. et al. On the reconstruction of functions from values at subsampled quadrature points, Math. Comput., 93 (346), 785–809 (2024).

24. Sickel, W., Ullrich, T. Smolyak's algorithm, sampling on sparse grids and function spaces of dominating mixed smoothness. Univ., 2006.