Geodesic
Reference points based transformation and approximation
Kybernetika, Tome 49 (2013) no. 4, pp. 644-662 Cet article a éte moissonné depuis la source Czech Digital Mathematics Library

Voir la notice de l'article

Interpolating and approximating polynomials have been living separately more than two centuries. Our aim is to propose a general parametric regression model that incorporates both interpolation and approximation. The paper introduces first a new $r$-point transformation that yields a function with a simpler geometrical structure than the original function. It uses $r \geq 2$ reference points and decreases the polynomial degree by $r-1$. Then a general representation of polynomials is proposed based on $r \geq 1$ reference points. The two-part model, which is suited to piecewise approximation, consist of an ordinary least squares polynomial regression and a reparameterized one. The later is the central component where the key role is played by the reference points. It is constructed based on the proposed representation of polynomials that is derived using the $r$-point transformation $T_r(x)$. The resulting polynomial passes through $r$ reference points and the other points approximates. Appropriately chosen reference points ensure quasi smooth transition between the two components and decrease the dimension of the LS normal matrix. We show that the model provides estimates with such statistical properties as consistency and asymptotic normality.
Interpolating and approximating polynomials have been living separately more than two centuries. Our aim is to propose a general parametric regression model that incorporates both interpolation and approximation. The paper introduces first a new $r$-point transformation that yields a function with a simpler geometrical structure than the original function. It uses $r \geq 2$ reference points and decreases the polynomial degree by $r-1$. Then a general representation of polynomials is proposed based on $r \geq 1$ reference points. The two-part model, which is suited to piecewise approximation, consist of an ordinary least squares polynomial regression and a reparameterized one. The later is the central component where the key role is played by the reference points. It is constructed based on the proposed representation of polynomials that is derived using the $r$-point transformation $T_r(x)$. The resulting polynomial passes through $r$ reference points and the other points approximates. Appropriately chosen reference points ensure quasi smooth transition between the two components and decrease the dimension of the LS normal matrix. We show that the model provides estimates with such statistical properties as consistency and asymptotic normality.
Classification : 41A10, 62F12, 62J05, 65D05, 65D07, 65D10
Keywords: polynomial representation; approximation model; smooth connection; consistency; asymptotic normality 
@article{KYB_2013_49_4_a9,
     author = {T\"or\"ok, Csaba},
     title = {Reference points based transformation and approximation},
     journal = {Kybernetika},
     pages = {644--662},
     year = {2013},
     volume = {49},
     number = {4},
     language = {en},
     url = {http://geodesic.mathdoc.fr/item/KYB_2013_49_4_a9/}
}
TY  - JOUR
AU  - Török, Csaba
TI  - Reference points based transformation and approximation
JO  - Kybernetika
PY  - 2013
SP  - 644
EP  - 662
VL  - 49
IS  - 4
UR  - http://geodesic.mathdoc.fr/item/KYB_2013_49_4_a9/
LA  - en
ID  - KYB_2013_49_4_a9
ER  - 
%0 Journal Article
%A Török, Csaba
%T Reference points based transformation and approximation
%J Kybernetika
%D 2013
%P 644-662
%V 49
%N 4
%U http://geodesic.mathdoc.fr/item/KYB_2013_49_4_a9/
%G en
%F KYB_2013_49_4_a9
Török, Csaba. Reference points based transformation and approximation. Kybernetika, Tome 49 (2013) no. 4, pp. 644-662. http://geodesic.mathdoc.fr/item/KYB_2013_49_4_a9/

[1] Csörgő, S., Mielniczuk, J.: Nonparametric regression under long-range dependent normal errors. Ann. Statist. 23 (1995), 3, 1000-1014. | DOI | MR | Zbl

[2] Dikoussar, N. D.: Function parametrization by using 4-point transforms. Comput. Phys. Commun. 99 (1997), 235-254. | DOI | Zbl

[3] Dikoussar, N. D.: Kusochno-kubicheskoje priblizhenije I sglazhivanije krivich v rezhime adaptacii. Communication JINR, P10-99-168, Dubna 1999.

[4] Dikusar, N. D.: The basic element method. Math. Models Comput. Simulat. 3 (2011), 4, 492-507. | DOI | MR | Zbl

[5] Dikoussar, N. D., Török, Cs.: Automatic knot finding for piecewise-cubic approximation. Matem. Mod. 18 (2006), 3, 23-40. | MR | Zbl

[6] Dikoussar, N. D., Török, Cs.: On one approach to local surface smoothing. Kybernetika 43 (2007), 4, 533-546. | MR | Zbl

[7] Eubank, R. L.: Nonparametric Regression and Spline Smoothing. Marcel Dekker, Inc., 1999. | MR | Zbl

[8] Kahaner, D., Moler, C., Nash, S.: Numerical Methods and Software. Prentice-Hall, Inc., 1989. | Zbl

[9] Kepič, T., Török, Cs., Dikoussar, N. D.: Wavelet compression. In: 13. International Workshop on Computational Statistics, Bratislava 2004, pp. 49-52.

[10] Matejčiková, A., Török, Cs.: Noise suppression in RDPT. Forum Statisticum Slovacum 3 (2005), 199-203.

[11] Nadaraya, E. A.: On estimating regression. Theory Probab. Appl. 9 (1964), 141-142. | Zbl

[12] Révayová, M., Török, Cs.: Piecewise approximation and neural networks. Kybernetika 43 (2007), 4, 547-559. | MR | Zbl

[13] Révayová, M., Török, Cs.: Reference points based recursive approximation. Kybernetika 49 (2013), 1, 60-72.

[14] Riplay, B. D.: Pattern Recognition and Neural Networks. Cambridge University Press 1996. | MR

[15] Seber, G. A. F.: Linear Regression Analysis. J. Wiley and Sons, New York 1977. | MR | Zbl

[16] Török, Cs.: 4-point transforms and approximation. Comput. Phys. Commun. 125 (2000), 154-166. | DOI | Zbl

[17] Török, Cs., Dikoussar, N. D.: Approximation with DPT. Comput. Math. Appl. 38 (1999), 211-220. | DOI | MR

[18] Trefethen, L. N.: Approximation Theory and Approximation Practice. SIAM, 2013. | MR | Zbl