Interpolation of functionals of stochastic sequences with stationary increments from observations with noise

2025

Read Interpolation of functionals of stochastic sequences with stationary increments from observations with noise (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-22764.txt.

Introduction

Noisy past and future data can help fill a gap in a sequence. Maksym Luz and Mykhailo Moklyachuk study that interpolation task in 2025. Their sequence has stationary increments. That means step changes have stable stats over time. A linear functional sums unknown middle values with weights.

Observations come from before the gap in clean form. After the gap, noise corrupts the sequence. The team derives the best linear estimate. They also give its mean square error. When spectra are known, the formulas are exact. When spectra are only known within a set, they use minimax rules. Those rules guard against the least favorable spectra.

Catalog record

YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-22764

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Interpolation of functionals of stochastic sequences with stationary increments from observations with noise Abstract Applied Statistics. Actuarial and Financial Mathematics. No. 2, 131–148, 2012 ISSN 1810-3022 Maksym Luz , Mykhailo Moklyachuk ∗ Department of Probability Theory, Statistics and Actuarial Mathematics, Taras Shevchenko National University of Kyiv, Kyiv 01601, Ukraine Corresponding email: [email protected] Abstract The problem of optimal estimation of linear functional {{A}_{N}}\xi=\sum\limits_{k=0}^{N}{a(k)\xi(k)}\, depending on the unknown values of a stochastic sequence \xi(m) with stationary n -th increments from observations of the sequence \xi(k) at points k=-1,-2,\ldots and of the sequence \xi(k)+\eta(k) at points of time k=N+1,N+2,\ldots is considered. Formulas for calculating the mean square error and the spectral characteristic of the optimal linear estimate of the functional are proposed under condition of spectral certainty, where spectral densities of the sequences \xi(m) and \eta(m) are exactly known. Minimax (robust) method of estimation is applied in the case where the spectral densities are not known exactly while some sets of admissible spectral densities are given. Formulas that determine the least favorable spectral densities and the minimax spectral characteristics are proposed for some specific sets of admissible densities Keywords sequence with stationary increments, minimax-robust estimation, mean square error, least favorable spectral density, minimax-robust spectral characteristic. AMS 2010 subject classifications. Primary: 60G10, 60G25, 60G35 Secondary: 62M20, 93E10 1 Introduction Stochastic sequences with stationary increments are an important generalization of wide sense stationary random sequences. The simplest examples of sequences with stationary increments are ARIMA sequences, periodic (seasonal) time series, etc. Random processes with stationary increments were studied by A.M. Yaglom in the article [1]. The article presented the spectral representation of stochastic increments, the canonical representation of the spectral density, and solution of the problem of prediction the value of the stochastic increment from known observations. The problem of prediction of processes with stationary increments is reduced to the problem of prediction the value of the stationary increment. The results of studies of random processes with stationary increments can be found in the works of M.S. Pinsker [2], A.M. Yaglom, and M.S. Pinsker [3]. In the work [4] A.M. Yaglom used the theory of generalized random processes to study processes with stationary increments. The classical methods of solution the problems of extrapolation, interpolation and filtering of stationary processes and sequences with known spectral densities were proposed by A.M. Kolmogorov [5], N. Wiener [6], A.M. Yaglom [7,8]. In the article by M.M. Luz [9], the problem of extrapolation of linear functionals from unknown values of a stationary process based on known observations at discrete moments of time is solved. In the case where the exact form of the spectral density of the process is unknown, while a set of admissible spectral densities is given, the minimax method of solution the problems of optimal estimation of unknown values of a stationary process is applied. The idea of the mentioned approach is that an estimate is determined that minimizes the value of the mean square error simultaneously for all densities from a set of admissible spectral densities. The minimax approach to the problem of extrapolation of stationary processes was first proposed by U. Grenander [10]. In the works of Y. Franke [11], Y. Franke and H. Poor [12], S. Kassam and H. Poor [13], the problems of minimax extrapolation and interpolation of stationary sequences are solved by convex optimization methods. In the works of M.P. Moklyachuk [14] – [18], the problems of extrapolation, interpolation and filtration for stationary processes and sequences are investigated. Vector stationary processes were investigated by Y.A. Rozanov [19]. In their paper O. Yu. Masyutka and M.P. Moklyachuk [20] obtained minimax estimates of linear functionals from unknown values of vector stationary processes and sequences. In the works by I.I. Dubovetska, M.P. Moklyachuk [21] and I.I. Dubovetska, O. Yu. Masyutka and M.P. Moklyachuk [22] the problems of minimax filtering and interpolation of periodically correlated sequences are solved. In the work of M.M. Luz and M.P. Moklyachuk [23], the problem of minimax interpolation for stochastic sequences with stationary increments for known values of this sequence is studied. This paper presented results of investigation the problem of optimal linear interpolation of the functional A_{N}\xi=\sum\limits_{k=0}^{N}{a(k)\xi(k)} from unknown values of a stochastic sequence with stationary n th increments \xi(m) based on observations of the sequence \xi(k) at points k=-1,-2,\ldots , and observations of the sequence \xi(k)+\eta(k) at points k=N+1,N+2,\ldots , where \eta(k) is a stochastic sequence with stationary n increments, uncorrelated with the sequence \xi(k) . The problem of interpolation of a sequence with stationary n increments is solved in the case where the spectral densities of the sequences \xi(m) and \eta(m) are known. In the case where the spectral densities are unknown while some admissible sets of spectral densities are given, the minimax estimation method is applied. For some given sets of admissible spectral densities, the least favorable densities and minimax spectral characteristics of the optimal estimate of the functional {{A}_{N}}\xi are determined. 2 Stationary increments. Spectral decomposition Definition 1 The stochastic n -th increment with step \mu\in Z of a stochastic sequence \{\xi(m),m\in Z\} is the function {{\xi}^{(n)}}(m,\mu)={{(1-{{B}_{\mu}})}^{n}}\xi(m)=\sum\limits_{l=0}^{n}{}\,{{(-1)}^{l}}C_{n}^{l}\xi(m-l\mu), (1) where {{B}_{\mu}} is the shift operator by \mu steps, such that {{B}_{\mu}}\xi(m)=\xi(m-\mu) , m\in Z . For the stochastic n -th increment, the relations hold true {{\xi}^{(n)}}(m,-\mu)={{(-1)}^{n}}{{\xi}^{(n)}}(m+n\mu,\mu), (2) {{\xi}^{(n)}}(m,k\mu)=\sum\limits_{l=0}^{(k-1)n}{{{A}_{l}}{{\xi}^{(n)}}(m-l\mu,\mu)}\,,\quad k\in N, (3) where \{{{A}_{l}},l=0,1,2,\ldots,(k-1)n\} are the coefficients at the corresponding powers of {{x}^{l}} of the polynomial expansion {{(1+x+\ldots+{{x}^{k-1}})}^{n}} in powers of x . Definition 2 The stochastic n -th increment {{\xi}^{(n)}}(m,\mu) with step \mu\in Z of a stochastic sequence \{\xi(m),m\in Z\} is called stationary (in the wide sense) if the mathematical expectations \mathbf{E}{{\xi}^{(n)}}({{m}_{0}},\mu)={{c}^{(n)}}(\mu),\quad\mathbf{E}{{\xi}^{(n)}}({{m}_{0}}+m,{{\mu}_{1}}){{\xi}^{(n)}}({{m}_{0}},{{\mu}_{2}})={{D}^{(n)}}(m,{{\mu}_{1}},{{\mu}_{2}}) exist for arbitrary integers {{m}_{0}}, m, \mu, {{\mu}_{1}}, {{\mu}_{2}} , and do not depend on {{m}_{0}} . The function {{c}^{(n)}}(\mu) is called the mean value of the stationary n -th increment, and the function {{D}^{(n)}}(m,{{\mu}_{1}},{{\mu}_{2}}) is called the structural function of the stationary n -th increment (or the structural function of the n -th order of the stochastic sequence \{\xi(m),m\in Z\} ). A stochastic sequence \{\xi(m),m\in Z\} that determines the stationary n -th increment {{\xi}^{(n)}}(m,\mu) by formula (1) is called a sequence with stationary n -th increments. Theorem 2.1 The mean value {{c}^{(n)}}(\mu) and the structure function {{D}^{(n)}}(m,{{\mu}_{1}},{{\mu}_{2}}) of the stationary stochastic n -th increment {{\xi}^{(n)}}(m,\mu) can be represented in the form {{c}^{(n)}}(\mu)=c{{\mu}^{n}}, (4) {{D}^{(n)}}(m,{{\mu}_{1}},{{\mu}_{2}})=\int\limits_{-\pi}^{\pi}{{{e}^{i\lambda m}}{{(1-{{e}^{-i{{\mu}_{1}}\lambda}})}^{n}}{{(1-{{e}^{i{{\mu}_{2}}\lambda}})}^{n}}\frac{1}{{{\lambda}^{2n}}}dF(\lambda)}\,, (5) where c is a constant, the function F(\lambda) is left continuous, nondecreasing, and bounded, F(-\pi)=0 . The constant c and the function F(\lambda) are uniquely determined by the increment {{\xi}^{(n)}}(m,\mu) . On the other hand, the function {{c}^{(n)}}(\mu) of the form (4) with a constant c and the function {{D}^{(n)}}(m,{{\mu}_{1}},{{\mu}_{2}}) of the form (5), where F(\lambda) satisfies the above conditions, are the mean value and the structure function of some stationary n -th increment {{\xi}^{(n)}}(m,\mu) . Using the representation (5) of the structure function of the stationary n -th increment {{\xi}^{(n)}}(m,\mu) and Karhunen’s theorem [24], [25], we obtain the representation of the stationary n -th increment {{\xi}^{(n)}}(m,\mu) in the following form {{\xi}^{(n)}}(m,\mu)=\int\limits_{-\pi}^{\pi}{{{e}^{im\lambda}}{{(1-{{e}^{-i\mu\lambda}})}^{n}}\frac{1}{{{(i\lambda)}^{n}}}dZ(\lambda)}\,, (6) where Z(\lambda) is an orthogonal stochastic measure on [-\pi,\pi) , which is subordinate to the spectral measure generated by the function F(\lambda) : \mathbf{E}Z({{A}_{1}})\overline{Z({{A}_{2}})}=F({{A}_{1}}\cap{{A}_{2}})<\infty. (7) We use the spectral decomposition (6) to find the optimal linear estimate of the unknown values of the stochastic sequence.

Read the complete text