Analysis of optimization methods for nonparametric estimation of probability density in large volume samples

A method is proposed for selecting the blurriness coefficient of kernel functions for nonparametric estimation of the probability density of a one-dimensional random variable with large volumes of statistical data, for example, obtained by remote sensing of natural objects. In the proposed method for selecting the blurriness coefficient, a regression estimate of the probability density is used. A method for synthesizing a regression probability density estimate is presented. The synthesis of the estimate is based on compression of the initial sample by decomposition of the range of values of a random variable. To decompose the range of values of a random variable, the Heinhold-Gaede rule and the formula for optimal selection of the number of sampling intervals are applied. Two approaches to the selection of the blurriness coefficient of the regression estimation of probability density using the traditional and proposed by the authors optimization methods of nonparametric estimation of probability density are considered. The traditional method of optimizing nonparametric estimation of probability density is based on minimizing its mean square deviation. In the proposed method, the selection of the blurriness coefficients of the kernel functions is based on the conditions of the minimum error of approximation of the regression estimate of the desired probability density. The approximation properties of the regression estimation of probability density using two methods of its optimization are analyzed. The conditions of their competence in estimating the probability densities of random variables with a lognormal distribution law are established. The results obtained allow for development when optimizing a regression estimate of the probability density of a multidimensional random variable.

1. Lapko A. V., Lapko V. A., Yadernye ocenki plotnosti veroyatnosti i ih primenenie [Kernel probability density estimates and their application], Krasnoyarsk, Reshetnev University Publ., 2021, 308 p. (In Russ.)

2. Lapko A. V., Lapko V. A., Optoelectronics, Instrumentation and Data Processing, 2014, vol. 50, no. 2, pp. 148–153. https://doi.org/10.3103/S875669901402006X

3. Rudemo M. Empirical choice of histogram and kernel density estimators, Scandinavian Journal of Statistics, 1982, no. 9, pp. 65–78.

4. Bowman A. W., Journal of Statistical Computation and Simulation, 1985, vol. 21, no. 3-4. https://doi.org/10.1080/00949658508810822

5. Hall P., Annals of Statistics, 1983, vol. 11(4), pp. 1156–1174. https://doi.org/10.1214/aos/1176346329

6. Jiang M., Provost S. B., Journal of Statistical Computation and Simulation, 2014, vol. 84, no. 3, pp. 614–627. https://doi.org/10.1080/00949655.2012.721366

7. Dutta S., Communications in Statistics – Simulation and Computation, 2016, vol. 45, no. 2, pp. 472–490. https://doi.org/10.1080/03610918.2013.862275

8. Sturges H. A., Journal of the American Statistical Association, 1926, vol. 21, pp. 65–66. https://doi.org/10.1080/01621459.1926.10502161

9. Storm R., Teoriya veroyatnostej. Matematicheskaya statistika. Statisticheskij kontrol’ kachestva [Probability theory. Mathematical statistics. Statistical quality control], Moscow, Mir Publ., 1970, 368 p. (In Russ.)

10. Heinhold I., Gaede K. W., Ingeniur statistic, München, Wien, Springler Verlag, 1964, 352 p. (In German)

11. Lapko A. V., Lapko V. A., Measurement Techniques, 2013, vol. 56, no. 7, pp. 763–767. https://doi.org/10.1007/s11018-013-0279-x

12. Lapko A. V., Lapko V. A., Measurement Techniques, 2020, vol. 63, no. 7, pp. 534–542. https://doi.org/10.1007/s11018-020-01820-1

13. Parzen E., Annals of Mathematical Statistics, 1962, vol. 33, nо. 3, pp. 1065–1076. https://doi.org/10.1214/aoms/1177704472

14. Epanechnikov V. A., Theory of Probability & Its Applications, 1969, vol. 14, no. 1, pp. 156–161. https://doi.org/10.1137/1114019

15. Gradov V. M., Ovechkin G. V., Ovechkin P. V., Rudakov I. V., Komp’yuternoe modelirovanie [Computer modeling], Moscow, Kurs, INFRA-M Publ., 2019, 264 p. (In Russ.)