zbMATH — the first resource for mathematics

An optimal experimental design for the Haar regression model. (English) Zbl 0814.62039
This paper presents the general framework for optimal design problems. Wavelets were introduced, and the Haar wavelet system was given as an example. For the Haar regression model of order \(m\), any design which concentrates mass \(2^{-(m+1)}\) in the \(2^{m+1}\) intervals \[ \{[ 2^{-(m+1)} i, \;2^{-(m+1)} (i+1)) \}_{i= 0,1,\dots, 2^{m+1} -1} \] is D-optimal. Because the Haar wavelet family is a set of orthonormal functions, the vector of estimated regression parameters is particularly simple to calculate. If the optimal design is used for the Haar wavelet system, and if the number of observations used is equal to the number of terms in the regression, then the vector of estimated regression parameters is equivalent to the discrete wavelet transform of the data.
As wavelets become more used, one will want to compare the classical designs for ordinary linear regression with those of Fourier regression and wavelet regression. The use of the Haar wavelet is the first step. In some ways, the simplest wavelet, the Haar wavelet, has the advantage of its rectangular shape, which enables one to make better comparisons even by eye.

62K05 Optimal statistical designs
33E15 Other wave functions
33E99 Other special functions
42C99 Nontrigonometric harmonic analysis
Full Text: DOI
[1] Benedetto, J.J., and Frazier, M.W., eds. (1994). Wavelets: Mathematics and Applications. CRC Press, Boca Raton, Fla. · Zbl 0840.00013
[2] Daubechies, Orthonormal bases of compactly supported wavelets, Comm. Pure Appl. Math. 41 pp 909– (1988) · Zbl 0644.42026
[3] Haar, Orthogonale Funktionensysteme, Math. Ann. 69 pp 331– (1910)
[4] Karlin, Optimal experimental designs, Ann. Math. Statist. 37 pp 783– (1966) · Zbl 0151.23904
[5] Karlin, Tchebycheff Systems: With Applications in Analysis and Statistics (1966) · Zbl 0153.38902
[6] Kiefer, The equivalence of two externum problems, Canad. J. Math. 12 pp 363– (1960) · Zbl 0093.15602 · doi:10.4153/CJM-1960-030-4
[7] Kupper, Minimax designs for Fourier series and spherical harmonics regressions: A characterization of rotatable arrangements, J. Roy. Statist. Soc. Ser. B 35 pp 493– (1972) · Zbl 0276.62071
[8] Laycock, Optimal design: Regression models for directions, Biometrika 62 pp 305– (1975)
[9] Mallat, A theory for multiresolution signal decomposition: The wavelet representation, IEEE Trans. Pattern Anal. and Machine Intelligence 11 pp 674– (1989) · Zbl 0709.94650
[10] Strang, Wavelets and dilation equations: A brief introduction, SIAM Rev. 31 pp 614– (1989) · Zbl 0683.42030
[11] Strang, Wavelet transforms vs. Fourier transforms, Bull. Amer. Math. Soc. (N.S.) 28 pp 288– (1993) · Zbl 0771.42021
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. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.