**(45-6) 17** * << * >> * Russian * English * Content * All Issues

I.V. Zenkov

660041, Krasnoyarsk, Russia, Svobodny Av. 79,

660036, Krasnoyarsk, Russia, Akademgorodok 50,

660037, Krasnoyarsk, Russia, Krasnoyarsky Rabochy Av. 31,

660049, Krasnoyarsk, Russia, Mira Av. 53

**DOI: **10.18287/2412-6179-CO-902

**Pages: **926-933.

**Full text of article:** Russian language.

**Abstract:**

We study in which way a priori information on the independence of random variables affects the approximation accuracy of a nonparametric estimate of the Rosenblatt–Parzen probability density. A new technique for generating sets of independent components of a multidimensional random variable is proposed. The methodology is based on testing the hypotheses of the independence of combinations of the multidimensional random variable components using a two-alternative nonparametric kernel algorithm for pattern recognition corresponding to the maximum likelihood criterion. Classes correspond to the domains of definition of the probability densities of sets of independent and dependent components of the multidimensional random variable. Nonparametric statistics of the kernel type are used to estimate the probability densities. The choice of the bandwidths of the kernel estimates of the probability densities is made from the condition of the minimum root-mean-square criterion. The sequential procedure for generating a set of independent components begins with the analysis of paired combinations of components of a multidimensional random variable. For each pair of components, the probability of an error in recognizing classes corresponding to the assumptions of independence and dependence of the considered components is estimated. A pair of components with the maximum difference between these errors is determined. If the errors obtained do not differ significantly, then there are no independent components in the considered multivariate random variable. If there is a significant difference in the probability estimates of class recognition errors, a pair of independent components is established. These components are included in a three-component set of a multidimensional random variable. The analysis of their combinations is carried out in the same way, following the above-described procedure. The process of generating the set of independent components is stopped when no reliable difference occurs any more between the probabilities of errors in recognizing situations belonging to the accepted classes. In this case, the previous set of independent components is the desired result. In contrast to the traditional methodology based on the Pearson criterion, the proposed approach allows us to bypass a problem of the decomposition of the range of values of random variables into multidimensional intervals.

The method of generating a set of independent components of a multidimensional random variable is illustrated by the results of the analysis of spectral features of remote sensing data of forest tracts using space imagery from the Landsat-8 satellite.

**Keywords**:

pattern recognition, information processing, optical data processing, hypothesis testing, forming a set of independent features, nonparametric pattern recognition algorithm, kernel probability density estimate, bandwidths selection of the kernel functions, remote sensing data.

**Citation**:

Zenkov IV, Lapko AV, Lapko VA., Kiryushina EV, Vokin VN, Bakhtina AV. A method of sequentially generating a set of components of a multidimensional random variable using a nonparametric pattern recognition algorithm. Computer Optics 2021; 45(6): 926-933. DOI: 10.18287/2412-6179-CO-902.

**Acknowledgements**:

The research was funded by the Russian Foundation for Basic Research, Government of the Krasnoyarsk Territory, and Krasnoyarsk Regional Science Foundation, project No. 20-41-240001.

References:

- Lapko AV, Lapko VA. Properties of nonparametric estimates of multidimensional probability density of independent random variables [In Russian]. Informatika i Sistemy Upravleniya 2012; 31(1): 166-174.

- Lapko AV, Lapko VA. Nonparametric estimation of probability density of independent random variables [In Russian]. Informatika i Sistemy Upravleniya 2011; 29(3): 118-124.

- Lapko AV, Lapko VA. Effect of a priori information about independence multidimensional random variables on the properties of their nonparametric density probability estimates [In Russian]. Sistemy Upravleniya i Informatsionnyye Tekhnologii 2012; 48(2.1): 164-167.

- Lapko AV, Lapko VA. Properties of the nonparametric decision function with a priori information on independence of attributes of classiﬁed objects. Optoelectronics, Instrumentation and Data Processing 2012; 48(4): 416-422. DOI: 10.3103/S8756699012040139.

- Pugachev VS. Probability theory and mathematical statistics: textbook [In Russian]. Moscow: “Fizmatlit” Publisher; 2002.

- Lapko AV, Lapko VA. Nonparametric algorithms of pattern recognition in the problem of testing a statistical hypothesis on identity of two distribution laws of random variables. Optoelectronics, Instrumentation and Data Processing 2010; 46(6): 545-550. DOI: 10.3103/S8756699011060069.

- Lapko AV, Lapko VA. Comparison of empirical and theoretical distribution functions of a random variable on the basis of a nonparametric classifier. Optoelectronics, Instrumentation and Data Processing 2012; 48(1): 37-41. DOI: 10.3103/S8756699012010050.

- Lapko AV, Lapko VA. A technique for testing hypotheses for distributions of multidimensional spectral data using a nonparametric pattern recognition algorithm. Computer Optics 2019; 43(2): 238-244. DOI: 10.18287/2412-6179-2019-43-2-238-244.

- Parzen E. On estimation of a probability density function and mode. Ann Math Statistic 1962; 33(3): 1065-1076. DOI: 10.1214/aoms/1177704472.

- Epanechnikov VA. Non-parametric estimation of a multivariate probability density. Theory Probab its Appl 1969; 14(1): 153-158. DOI: 10.1137/1114019.

- Rudemo M. Empirical choice of histogram and kernel density estimators. Scand Stat Theory Appl 1982; 9(2): 65-78.

- Hall P. Large-sample optimality of least squares cross-validation in density estimation. Annals of Statistics 1983; 11(4): 1156-1174. DOI: 10.1214/aos/1176346329.

- Silverman BW. Density estimation for statistics and data analysis. London: Chapman and Hall; 1986.

- Sheather S, Jones M. A reliable data-based bandwidth selection method for kernel density estimation. J R Stat Soc Series B 1991; 53(3): 683-690. DOI: 10.1111/j.2517-6161.1991.tb01857.x.

- Sheather SJ. Density estimation. Stat Sci 2004; 19(4): 588-597. DOI: 10.1214/088342304000000297.

- Terrell GR, Scott DW. Oversmoothed nonparametric density estimates. J Am Stat Assoc 1985; 80(389): 209-214. DOI: 10.1080/01621459.1985.10477163.

- Jones MC, Marron JS, Sheather SJ. A brief survey of bandwidth selection for density estimation. J Am Stat Assoc 1996; 91(433): 401-407. DOI: 10.2307/2291420.

- Scott DW. Multivariate density estimation: Theory, practice, and visualization. New Jersey: John Wiley and Sons; 2015.

- Lapko AV, Lapko VA. Modified algorithm for rapid choice of spread coefficients for kernel estimates of multidimensional probability densities. Measurement Techniques 2021; 63(11): 856-861. DOI: 10.1007/s11018-021-01873-w.

- Lapko AV, Lapko VA. Analysis of optimization methods for nonparametric estimation of the probability density with respect to the blur factor of kernel functions. Measurement Techniques 2017; 60(6): 515-522. DOI: 10.1007/s11018-017-1228-x.
- Sharakshaneh АS, Zheleznov IG, Ivnitskij VА. Complex system [In Russian]. Moscow: “Vysshaya shkola” Publisher; 1977.

© 2009, IPSI RAS151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20