×

Tikhonov regularization for polynomial approximation problems in Gauss quadrature points. (English) Zbl 1461.65010

Summary: This paper is concerned with the introduction of Tikhonov regularization into least squares approximation scheme on \([-1, 1]\) by orthonormal polynomials, in order to handle noisy data. This scheme includes interpolation and hyperinterpolation as special cases. With Gauss quadrature points employed as nodes, coefficients of the approximation polynomial with respect to given basis are derived in an entry-wise closed form. Under interpolatory conditions, the solution to the regularized approximation problem is rewritten in forms of two kinds of barycentric interpolation formulae, by introducing only a multiplicative correction factor into both classical barycentric formulae. An \(L_2\) error bound and a uniform error bound are derived, providing similar information that Tikhonov regularization is able to reduce the operator norm (Lebesgue constant) and the error term related to the level of noise, both by multiplying a correction factor which is less than one. Numerical examples show the benefits of Tikhonov regularization when data is noisy or data size is relatively small.

MSC:

65D05 Numerical interpolation
65D30 Numerical integration
41A10 Approximation by polynomials

Software:

Chebfun; OPQ
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] Berrut J-P and Trefethen L N 2004 Barycentric Lagrange interpolation SIAM Rev.46 501-17 · Zbl 1061.65006 · doi:10.1137/s0036144502417715
[2] Driscoll T A, Hale N and Trefethen L N 2014 Chebfun Guide (Oxford: Pafnuty Publications)
[3] Filbir F and Mhaskar H N 2011 Marcinkiewicz-Zygmund measures on manifolds J. Complexity27 568-96 · Zbl 1235.58007 · doi:10.1016/j.jco.2011.03.002
[4] Gautschi W 2004 Orthogonal Polynomials(Computation and Approximation) (Oxford: Oxford University Press) · Zbl 1130.42300 · doi:10.1093/oso/9780198506720.001.0001
[5] Gautschi W 2012 Numerical Analysis 2nd edn (Basel: Birkhäuser) · Zbl 1378.65002 · doi:10.1007/978-0-8176-8259-0
[6] Glaser A, Liu X and Rokhlin V 2007 A fast algorithm for the calculation of the roots of special functions SIAM J. Sci. Comput.29 1420-38 · Zbl 1145.65015 · doi:10.1137/06067016x
[7] Hesse K, Sloan I H and Womersley R S 2017 Radial basis function approximation of noisy scattered data on the sphere Numer. Math.137 579-605 · Zbl 1380.41005 · doi:10.1007/s00211-017-0886-6
[8] Higham N J 2004 The numerical stability of barycentric Lagrange interpolation IMA J. Numer. Anal.24 547-56 · Zbl 1067.65016 · doi:10.1093/imanum/24.4.547
[9] Kress R 1998 Numerical Analysis(Graduate Texts in Mathematics vol 181) (Berlin: Springer) · Zbl 0913.65001 · doi:10.1007/978-1-4612-0599-9
[10] Lazarov R D, Lu S and Pereverzev S V 2007 On the balancing principle for some problems of numerical analysis Numer. Math.106 659-89 · Zbl 1129.65041 · doi:10.1007/s00211-007-0076-z
[11] Le Gia Q T and Mhaskar H N 2008 Localized linear polynomial operators and quadrature formulas on the sphere SIAM J. Numer. Anal.47 440-66 · Zbl 1190.65039 · doi:10.1137/060678555
[12] Lu S, Naumova V and Pereverzev S V 2013 Legendre polynomials as a recommended basis for numerical differentiation in the presence of stochastic white noise J. Inverse Ill-Posed Problems21 193-216 · Zbl 1276.65016 · doi:10.1515/jip-2012-0050
[13] Lu S and Pereverzev S V 2009 Sparse recovery by the standard Tikhonov method Numer. Math.112 403-24 · Zbl 1167.65025 · doi:10.1007/s00211-009-0214-x
[14] Lu S and Pereverzev S V 2013 Regularization Theory for Ill-Posed Problems(Selected Topics) vol 58 (Berlin: de Gruyter & Co) · Zbl 1282.47001 · doi:10.1515/9783110286496
[15] Lu S, Pereverzev S V and Ramlau R 2006 An analysis of Tikhonov regularization for nonlinear ill-posed problems under a general smoothness assumption Inverse Problems23 217 · Zbl 1118.65056 · doi:10.1088/0266-5611/23/1/011
[16] Lu S, Pereverzev S V, Shao Y and Tautenhahn U 2010 Discrepancy curves for multi-parameter regularization J. Inverse Ill-Posed Problems18 655-76 · Zbl 1280.47016 · doi:10.1515/jiip.2010.030
[17] Lu S, Pereverzev S V and Tautenhahn U 2010 A model function method in regularized total least squares Appl. Anal.89 1693-703 · Zbl 1201.65096 · doi:10.1080/00036811.2010.492502
[18] Lu S, Pereverzev S V and Tautenhahn U 2010 Regularized total least squares: computational aspects and error bounds SIAM J. Matrix Anal. Appl.31 918-41 · Zbl 1198.65094 · doi:10.1137/070709086
[19] Mhaskar H N 2004 Polynomial operators and local smoothness classes on the unit interval J. Approx. Theory131 243-67 · Zbl 1083.41004 · doi:10.1016/j.jat.2004.10.002
[20] Mhaskar H N 2005 Polynomial operators and local smoothness classes on the unit interval, II Jaen J. Approx. Theor.1 1-25 · Zbl 1181.41047
[21] Mhaskar H N 2020 A direct approach for function approximation on data defined manifolds Neural Netw.132 253-68 · Zbl 1475.68319 · doi:10.1016/j.neunet.2020.08.018
[22] Mhaskar H N, Narcowich F J and Ward J D 2001 Spherical Marcinkiewicz-Zygmund inequalities and positive quadrature Math. Comput.70 1113-30 · Zbl 0980.76070 · doi:10.1090/S0025-5718-00-01240-0
[23] Mhaskar H N, Naumova V and Pereverzyev S V 2013 Filtered Legendre expansion method for numerical differentiation at the boundary point with application to blood glucose predictions Appl. Math. Comput.224 835-47 · Zbl 1334.65058 · doi:10.1016/j.amc.2013.09.015
[24] Mhaskar H N, Pereverzyev S V and van der Walt M D 2017 A deep learning approach to diabetic blood glucose prediction Front. Appl. Math. Stat.3 14 · doi:10.3389/fams.2017.00014
[25] Pereverzyev S V, Sloan I H and Tkachenko P 2015 Parameter choice strategies for least-squares approximation of noisy smooth functions on the sphere SIAM J. Numer. Anal.53 820-35 · Zbl 1314.65040 · doi:10.1137/140964990
[26] Rivlin T J 1969 An Introduction to the Approximation of Functions (Waltham: Blaisdell) · Zbl 0189.06601
[27] Rutishauser H 1990 Lectures on Numerical Mathematics (Basel: Birkhäuser) · Zbl 0699.65002 · doi:10.1007/978-1-4612-3468-5
[28] Salzer H E 1972 Lagrangian interpolation at the Chebyshev points xn,ν ≡ cos(νπ/n), ν = O(1)n; some unnoted advantages Comput. J.15 156-9 · Zbl 0242.65007 · doi:10.1093/comjnl/15.2.156
[29] Schwarz H R and Waldvogel J 1989 Numerical Analysis(A Comprehensive Introduction) (New York: Wiley) · Zbl 0715.65003
[30] Sloan I H 1995 Polynomial interpolation and hyperinterpolation over general regions J. Approx. Theor.83 238-54 · Zbl 0839.41006 · doi:10.1006/jath.1995.1119
[31] Szegő G 1939 Orthogonal Polynomials(Colloquium Publications vol 23) (Providence, RI: American Mathematical Society) · JFM 65.0286.02 · doi:10.1090/coll/023
[32] Tikhonov A N and Arsenin V J 1977 Solutions of Ill-Posed Problems (Washington, D.C: Winston & Sons) · Zbl 0354.65028
[33] Trefethen L N 2013 Approximation Theory and Approximation Practice vol 128 (Philadelphia, PA: SIAM) · Zbl 1264.41001
[34] Wang H, Huybrechs D and Vandewalle S 2014 Explicit barycentric weights for polynomial interpolation in the roots or extrema of classical orthogonal polynomials Math. Comput.83 2893-914 · Zbl 1297.41001 · doi:10.1090/s0025-5718-2014-02821-4
[35] Wang H and Xiang S 2012 On the convergence rates of Legendre approximation Math. Comput.81 861-77 · Zbl 1242.41016 · doi:10.1090/s0025-5718-2011-02549-4
[36] Wei Y, Xie P and Zhang L 2016 Tikhonov regularization and randomized GSVD SIAM J. Matrix Anal. Appl.37 649-75 · Zbl 1339.65057 · doi:10.1137/15m1030200
[37] Xiang H and Zou J 2013 Regularization with randomized SVD for large-scale discrete inverse problems Inverse Problems29 085008 · Zbl 1286.65053 · doi:10.1088/0266-5611/29/8/085008
[38] Xiang H and Zou J 2015 Randomized algorithms for large-scale inverse problems with general Tikhonov regularizations Inverse Problems31 24 · Zbl 1327.65077 · doi:10.1088/0266-5611/31/8/085008
[39] Xiang S 2012 On error bounds for orthogonal polynomial expansions and Gauss-type quadrature SIAM J. Numer. Anal.50 1240-63 · Zbl 1263.65027 · doi:10.1137/110820841
[40] Zhong M, Lu S and Cheng J 2012 Multiscale analysis for ill-posed problems with semi-discrete Tikhonov regularization Inverse Problems28 065019 · Zbl 1254.65071 · doi:10.1088/0266-5611/28/6/065019
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.