GUARANTEED ROOT MEAN SQUARE ESTIMATES OF OBSERVATIONS WITH UNKNOWN MATRICES

  • O. G. Nakonechnyi Faculty of Computer Science and Cybernetics, Taras Shevchenko National University of Kyiv, Kyiv, Ukraine
  • G. I. Kudin Faculty of Computer Science and Cybernetics, Taras Shevchenko National University of Kyiv, Kyiv, Ukraine
  • P. M. Zinko Faculty of Computer Science and Cybernetics, Taras Shevchenko National University of Kyiv, Kyiv, Ukraine
  • T. P. Zinko Faculty of Computer Science and Cybernetics, Taras Shevchenko National University of Kyiv, Kyiv, Ukraine
Keywords: linear estimation, rms guaranteed estimate, estimation error, quasi-minimax guaranteed estimate, perturbed observation matrix, correlation matrix, operator equation, small parameter, statistical uncertainty

Abstract

The problems of guaranteed mean square estimation of unknown rectangular matrices based on observations of linear functions from random matrices with random errors are considered in the paper. Asymptotic distributions of guaranteed errors and guaranteed estimates are obtained in the case of small perturbations of the matrices. A test example of the asymptotic distribution is given.

References

Chatterjee S. Matrix Estimation by Universal Singular Value Thresholding. The Annals of Statistics. 2015. Vol. 43. No 1. P. 177–214.

Negahban S., Wainwright M. J. Estimation of (near) low-rank matrices with noise and high-dimensional scaling. Ann. Statist. 2011, 39. P. 1069–1097. MR2816348

Albert A. Regression, Pseudoinverse, Recurrent Estimation (translated from English). Moscow: Nauka, 1977. 305 p.

Arnold B. F., Stanlecker P. Linear estimation in regression analysis using fuzzy prior information. Rardom Oper. And Stoch. Equ. 1997, Vol. 5, No 2, P. 105–116.

Nakonechnyi O., Michalek J. Minimax estimates of a linear parameter function in a regression model under restrictions on the parameters and variance-covariance matrix. Journal of Numerical and Applied Mathematics. 2000, No 1 (102). P. 3790–3802.

Christopeit N., Helmes K. Linear minimax estimation with ellipsoidal constraints. Acta Applicandae Mathematicae. 1996, 43, I1, P. 3–15.

Girko V. Spectral Theory of Minimax Estimation. Acta Applicandae Mathematicae. 1996. 43, 1, P. 59–69.

Trenkler G. Quasi minimax estimation in the Linear regression model. Statistics. 18. 1987. P. 219–226.

Draper N., Smith H. Applied Regression Analysis. Publishing house “Williams”, 2007. 912 p.

Nakonechnyi O. G., Kudin G. I., Zinko T. P. Formulas of perturbation for one class of inverse operators. Matematychni Studii. 2019. 52, No 2. P. 124–132.

Nakonechnyi O. G., Kudin G. I., Zinko P. M., Zinko T. P. Perturbation Method in Problems of Linear Matrix Regression. Journal of Automation and Information Sciences. 2020. No 1. P. 38–47.

Nakonechnyi O. G., Kudin G. I., Zinko P. M., Zinko T. P. Approximate guaranteed estimates of matrices in linear regression problems with a small parameter. System research and information technologies. 2020. No 4. P. 88–102.

Nakonechnyi O. G., Kudin G. I., Zinko P. M., Zinko T. P. Guaranteed root-mean-square estimates of linear transformations of matrices in the conditional of statistical uncertainty. Journal of Automation and Information Sciences. 2021. No 2. P. 24–37.

Nakonechnyi O. G., Kudin G. I., Zinko P. M., Zinko T. P. Minimax root-mean-square estimates of matrix parameters in linear regression problems under uncertainty. Journal of Automation and Information Sciences. 2021. No 4. P. 28–37.

Published
2023-02-01
How to Cite
Nakonechnyi, O., Kudin, G., Zinko, P., & Zinko, T. (2023). GUARANTEED ROOT MEAN SQUARE ESTIMATES OF OBSERVATIONS WITH UNKNOWN MATRICES. Journal of Numerical and Applied Mathematics, 1(2), 98-115. https://doi.org/10.17721/2706-9699.2022.2.12