Abstract
In this paper, we develop an efficient nonparametric estimation theory for continuous time regression models with non-Gaussian Lévy noises in the case when the unknown functions belong to Sobolev ellipses. Using the Pinsker’s approach, we provide a sharp lower bound for the normalized asymptotic mean square accuracy. However, the main result obtained by Pinsker for the Gaussian white noise model is not correct without additional conditions for the ellipse coefficients. We find such constructive sufficient conditions under which we develop efficient estimation methods. We show that the obtained conditions hold for the ellipse coefficients of an exponential form. For exponential coefficients, the sharp lower bound is calculated in explicit form. Finally, we apply this result to signals number detection problems in multi-pass connection channels and we obtain an almost parametric convergence rate that is natural for this case, which significantly improves the rate with respect to power-form coefficients.
Similar content being viewed by others
References
Beltaief, S., Chernoyarov, O., Pergamenshchikov, S. M. (2020). Model selection for the robust efficient signal processing observed with small Levy noise. Annals of the Institute of Statistical Mathematics, 72, 1205–1235.
Hodara, P., Krell, N., Löcherbach, E. (2018). Non-parametric estimation of the spiking rate in systems of interacting neurons. Statistical Inference for Stochastic Processes, 21, 81–111.
Ibragimov, I. A., Khasminskii, R. Z. (1981). Statistical estimation: Asymptotic theory. New York: Springer.
Kassam, S. A. (1988). Signal detection in non-Gaussian noise. IX. New York: Springer.
Konev, V. V., Pergamenshchikov, S. M. (2009a). Nonparametric estimation in a semimartingale regression model. Part 1. Oracle Inequalities. Vestnik Tomskogo Gosudarstvennogo Universiteta. Matematika i Mekhanika, 3(7), 23–41.
Konev, V. V., Pergamenshchikov, S. M. (2009b). Nonparametric estimation in a semimartingale regression model. Part 2. Robust asymptotic efficiency. Vestnik Tomskogo Gosudarstvennogo Universiteta. Matematika i Mekhanika, 4(8), 31–45.
Konev, V. V., Pergamenshchikov, S. M. (2012). Efficient robust nonparametric in a semimartingale regression model. Annales de lInstitut Henri Poincare, 48(4), 1217–1244.
Konev, V. V., Pergamenshchikov, S. M. (2015). Robust model selection for a semimartingale continuous time regression from discrete data. Stochastic Processes and Their Applications, 125, 294–326.
Konev, V. V., Pergamenshchikov, S. M., Pchelintsev, E. A. (2014). Estimation of a regression with the pulse type noise from discrete data. Theory of Probability & its Applications, 58(3), 442–457.
Kuks, I. A., Olman, V. (1971). A minimax linear estimator of regression coefficients. Izv. Akad. Nauk Eston. SSR, 20, 480–482.
Kutoyants, Yu. A. (1984). Parameter Estimation for Stochastic Processes. Berlin: Heldeman-Verlag.
Kutoyants, Yu. A. (1994). Identification of dynamical systems with small noise. Dordrecht: Kluwer Academic Publishers Group.
Le Cam, L. (1990). Asymptotic methods in statistical decision theory. Springer series in statistics. New York: Springer.
Lepski, O. V. (1990). One problem of adaptive estimation in Gaussian white noise. Theory of Probability & Its Applications, 35, 459–470.
Lepski, O. V., Spokoiny, V. G. (1997). Optimal pointwise adaptive methods in nonparametric estimation. The Annals of Statistics, 25, 2512–2546.
Liptser, R. S., Shiryaev, A. N. (1989). Theory of martingales. New York: Springer.
Nemirovskii, A. (2000). Topics in non-parametric statistics. Lecture Notes in Mathematics, 1738, 85–277.
Pchelintsev, E. (2013). Improved estimation in a non-Gaussian parametric regression. Statistical Inference for Stochastic Processes, 16(1), 15–28.
Pchelintsev, E., Pergamenshchikov, S. (2018). Oracle inequalities for the stochastic differential equations. Statistical Inference for Stochastic Processes, 21(2), 469–483.
Pchelintsev, E., Pergamenshchikov, S. (2019). Adaptive model selection method for a conditionally Gaussian semimartingale regression in continuous time. Vestnik Tomskogo Gosudarstvennogo Universiteta. Matematika i Mekhanika, 58, 14–31.
Pchelintsev, E. A., Pchelintsev, V. A., Pergamenshchikov, S. M. (2018). Non asymptotic sharp oracle inequality for the improved model selection procedures for the adaptive nonparametric signal estimation problem. Communications—Scientific Letters of the University of Zilina, 20(1), 72–76.
Pchelintsev, E. A., Pchelintsev, V. A., Pergamenshchikov, S. M. (2019). Improved robust model selection methods for a Lévy nonparametric regression in continuous time. Journal of Nonparametric Statistics, 31(3), 612–628.
Pinsker, M. S. (1981). Optimal filtration of square integrable signals in gaussian white noise. Problems of Transimission Information, 17, 120–133.
Tsybakov, A. B. (1998). Pointwise and sup-norm sharp adaptive estimation of functions on the Sobolev classes. Annals of Statistics, 26, 2420–2469.
Tsybakov, A. B. (2009). Introduction in nonparametric estimation. New York: Springer.
Acknowledgements
The authors are grateful to the anonymous referees and to the AE for careful reading and for helpful comments. The results of Sect. 6 have been obtained under Grant of the President of the Russian Federation no. MK-834.2020.9.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was supported by RSF, Grant no 20-61-47043.
Auxiliary results
Auxiliary results
1.1 Proof of Lemma 1
First, note that the mean square error (13) can be represented as
Recall that
In view of (10) and \(\mathbf{E}_{\theta }\xi _{j}=0\), one has
and
Furthermore, using (73) and (72), we can rewrite
and, therefore,
Note that \(\gamma _{j}\) from (73) cannot be used in (12), because they depend of unknown parameters, and the estimate \(\widehat{S}_{\gamma }(t)\) cannot be calculated. By the Lagrange method, we obtain that
where \((x)_{+}=\max (0,x)\) and the Lagrange coefficient \(\mu\) is the solution of the following equation
If the condition \(\mathbf{A}_{1})\) holds, then the function \(f(\mu )\) is continuous increasing function with \(f(0)=0\) and \(\lim _{\mu \rightarrow \infty }f(\mu )=\infty\), and we can deduce that this equation has an unique solution
where \(\mathbf{m}=N_{\mu ^{2}}\) is defined in (7). This implies that
Setting \(g(n)=\sqrt{a_n}\sum _{j=1}^{n}\sqrt{a_{j}}-\sum _{j=1}^{n}a_{j}\) and using the definition (74), we obtain that
Therefore, in view of the definition (14), we find
Hence, Lemma 1. \(\Box\)
1.2 Representation for the \(\sigma\)-field generated by \((\xi _{t})_{0\le t\le 1}\).
Lemma 3
Let \((\varphi _{k})_{k\ge 1}\) be arbitrary orthonormal basis in \({{\mathcal {L}}}_{2}[0,1]\) with \(\varphi _{1}\equiv 1\). Then,
where \(\xi _{k}=\int ^{1}_{0}\,\varphi _{k}(t)\mathrm {d}\xi _{t}\).
Proof
Let \((\text{ Tr}_{j})_{j\ge 1}\) be the trigonometric basis in \({{\mathcal {L}}}_{2}[0,1]\). Taking into account that any trajectory of the process \(\xi\) belongs to \({{\mathcal {L}}}_{2}[0,1]\), we can represent it as
Using here the Ito formula, we obtain that
Note now that the functions \(\widetilde{\text{ Tr }}_{j}\) can be represented as
In view of \(\xi _{1}=\int ^{1}_{0}\varphi _{1}(s)\mathrm {d}\xi _{s}\), we can rewrite the coefficients \(\tau _{j}\) as
So, the coefficients \(\tau _{j}\) are measurable with respect to the \(\sigma\)-field \(\sigma \{\xi _{k}\,,\,k\ge 1\}\), and therefore, the Brownian motion is measurable with respect to this \(\sigma\)-field also, i.e. \(\sigma \{\xi _{t}\,,\,0\le t\le 1\}\subseteq \sigma \{\xi _{k}\,,\,k\ge 1\}\). The inverse inclusion is obvious. Hence, Lemma 3. \(\square\)
1.3 Conditional distribution tool
Lemma 4
Let \(\zeta\) and \(\xi\) be independent Gaussian random variables with the parameters \((0,\theta ^2)\) and \((\nu ,\sigma ^{2})\) , respectively, and let \(\eta =\zeta +\xi\). Then,
Proof
Note that \(\eta\) is Gaussian random variable with the parameters \((\nu ,\theta ^2+\sigma ^2)\). By the definition of the conditional expectation
and \(\mathbf{p}_{\zeta |\eta }(x|y)\) is the corresponding conditional distribution density
where
This implies that
Hence Lemma 4. \(\square\)
Lemma 5
Let \(\zeta\) and \(\xi\) be two independent random variables such that \(\zeta\) is uniformly distributed on \((-\theta ,\theta )\) for some \(\theta >0\) and \(\xi\) is Gaussian with the parameters \((\nu ,\sigma ^{2})\), and let \(\eta =\zeta +\xi\). Then,
where \(L=\theta /\sigma\).
Proof
First, note that
where \(\mathbf{m}(\eta )=\mathbf{E}({\tilde{\xi }}|\eta )\) and \({\tilde{\xi }}=\xi -\nu\). It is clear that
where \(\mathbf{p}_{\eta |{\tilde{\xi }}}(z|x)\), \(\mathbf{p}_{{\tilde{\xi }}}\) and \(\mathbf{p}_{\eta }\) are the corresponding distribution densities. Since the random variables \(\zeta\) and \({\tilde{\xi }}\) are independent and \(\zeta\) is uniform on the interval \((-\theta ,\theta )\), then
Here \(\mathbf{p}_{{\tilde{\xi }}}(x)=\sigma ^{-1}\phi (x/\sigma )\), where \(\phi\) is the (0, 1)-Gaussian density. Now, we have
where
For \(|{\tilde{z}}|<(1-\epsilon )\theta\) with \(\epsilon =1/\sqrt{L}\), the indicator \(\mathbf{1}_{\varGamma }\rightarrow 1\) as \(L\rightarrow \infty\) and, therefore, \(m(z)/\sigma \rightarrow 0\). Let now \(\rho _{L}=m^2(\eta )/ \sigma ^2=(\mathbf{E}({\overline{\xi }}|\eta ))^{2}\) and \({\overline{\xi }}=\widetilde{\xi }/\sigma \sim {{\mathcal {N}}}(0,1)\). Then,
where \(\widetilde{\eta }=\eta -\nu\). By the Jensen inequality \(\rho _{L}^{2} \le \mathbf{E}\left( {\overline{\xi }}^{4}\, \mid \, \eta \right)\), and, therefore,
i.e. \((\rho _{L})_{L\ge 1}\) is uniformly integrable. Since the random variables \(\rho _{L}\mathbf{1}_{\left\langle |\widetilde{\eta }| <(1-\epsilon )\theta \right\rangle }\rightarrow 0\) as \(L\rightarrow \infty\) almost sure, then \(\mathbf{E}\rho _{L}\mathbf{1}_{\left\langle |\widetilde{\eta }| <(1-\epsilon )\theta \right\rangle }\rightarrow 0\) as \(L\rightarrow \infty\). Moreover, taking into account that \(\mathbf{p}_{\eta }(z)\le 1/2\theta\), we get
Further, we have
Hence, Lemma 5. \(\square\)
Lemma 6
Let \(\zeta\) and \(\xi\) be two independent random variables, such that \(\mathbf{P}(\zeta =-\theta )=\mathbf{P}(\zeta =\theta )=1/2\) for some \(\theta >0\) and \(\xi\) is Gaussian with the parameters \((\nu ,\sigma ^{2})\), and let \(\eta =\zeta +\xi\). Then
where \(L=\theta /\sigma\).
Proof
First, note that in this case
where \(\widetilde{\eta }=(\eta -\nu )/\sigma\) and \(\phi\) is the (0, 1)-Gaussian density. It is clear that \(\vert \rho _{L}(x)\vert \le 1\) and
Therefore,
Taking into account here that \(\mathbf{E}\,\widetilde{\eta }^{2}\le 2L^{2}+2\) and passing to the limit as \(\lim _{M\rightarrow \infty }\lim _{L\rightarrow 0}\), we obtain (75). Hence, Lemma 6. \(\square\)
About this article
Cite this article
Pchelintsev, E., Pergamenshchikov, S. & Povzun, M. Efficient estimation methods for non-Gaussian regression models in continuous time. Ann Inst Stat Math 74, 113–142 (2022). https://doi.org/10.1007/s10463-021-00790-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10463-021-00790-7