Orthogonal Polynomial Density Estimates: Alternative Representation and Degree Selection

Authors: Serge B. Provost, Min Jiang

Abstract:

The density estimates considered in this paper comprise a base density and an adjustment component consisting of a linear combination of orthogonal polynomials. It is shown that, in the context of density approximation, the coefficients of the linear combination can be determined either from a moment-matching technique or a weighted least-squares approach. A kernel representation of the corresponding density estimates is obtained. Additionally, two refinements of the Kronmal-Tarter stopping criterion are proposed for determining the degree of the polynomial adjustment. By way of illustration, the density estimation methodology advocated herein is applied to two data sets.

Keywords: kernel density estimation, orthogonal polynomials, moment-based methodologies, density approximation.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1329791

References:

[1] Izenman A J. Recent development in nonparametric density density estimation. Journal of the American Statistical Association, 1991, 86:205- 224.
[2] Watson G S. Density estimation by orthogonal series. Annals of Mathematical Statistics, 1969, 40:1496-1498.
[3] Rosenblatt M. Curve estimates. The Annals of Mathematical Statistics, 1971, 42:1815-1842.
[4] Hall P. On the rate of convergence of orthogonal series density estimators. Journal of Royal the Royal Statistical Society, Ser. B, 1986, 48:115-122.
[5] Johnstone I M, Silverman B W. Speed of estimation in positron emission tomography and related inverse problems. The Annals of Statistics, 1990, 18:251-280.
[6] Provost S B. Moment-based density approximants. The Mathematica Journal, 2005, 9:727-756.
[7] Ha H-T, Provost S B. A viable alternative to resorting to statistical tables. Communications in Statistics-Simulation and Computation, 2007, 36:1135-1151.
[8] Provost S B, Ha H-T. On the inversion of certain moment matrices. Linear Algebra and Its Applications, 2009, 430:2650-2658.
[9] Brunk H B. Univariate density estimation by orthogonal series. Biometrika, 1978, 65:521-528.
[10] Parzen E. Nonparametric statistical data modeling. Journal of the American Statistical Association, 1979, 74:105-121.
[11] Ruppert D, Cline D B H. Bias reduction in kernel density estimation by smoothed empirical transformations. Institute of Mathematical Statistics, 1994, 22:185-210.
[12] Hjort N L, Jones M C. Locally parametric nonparametric density estimation. The Annals of Statistics, 1996, 24:1619-1647.
[13] Rao C R. Linear Statistical Inference and Its Applications. New York: Wiley, 1973.
[14] Arfken G. Gram-Schmidt Orthogonalization. Orlando: Academic Press, 1985.
[15] Anderson G L, De Figueiredo R J P. An adaptive orthogonal-series estimator for probability density functions. The Annals of Statistics, 1980, 8:347-376.
[16] Alexits G. Convergence Problems of Orthogonal Series. New York: Pergamon Press, 1961.
[17] Hildebrand F B. Introduction to Numerical Analysis. New York: McGraw-Hill, 1956.
[18] Kronmal R, Tarter M. The estimation of probability densities and cumulatives by Fourier series methods. Journal of the American Statistical Association, 1968, 63:925-952.
[19] Diggle P J, Hall P. The selection of terms in an orthogonal series density estimator. Journal of American Statistical Association, 1986, 81:230-233.
[20] Tarter M, Kronmal R. An introduction to the implementation and theory of nonparametric density estimation. The American Statistician, 1976, 30:105-112.
[21] Simonoff J S. Smoothing Methods in Statistics. New York: Springer, 1996.
[22] Roeder K. Density estimation with confidence sets exemplified by superclusters and voids in galaxies. Journal of the American Statistical Association, 1990, 85:617-624.