Annals of Statistics

Adaptive invariant density estimation for ergodic diffusions over anisotropic classes

Claudia Strauch

Abstract

Consider some multivariate diffusion process $\mathbf{X}=(X_{t})_{t\geq0}$ with unique invariant probability measure and associated invariant density $\rho$, and assume that a continuous record of observations $X^{T}=(X_{t})_{0\leq t\leq T}$ of $\mathbf{X}$ is available. Recent results on functional inequalities for symmetric Markov semigroups are used in the statistical analysis of kernel estimators $\widehat{\rho}_{T}=\widehat{\rho}_{T}(X^{T})$ of $\rho$. For the basic problem of estimation with respect to $\mathrm{sup}$-norm risk under anisotropic Hölder smoothness constraints, the proposed approach yields an adaptive estimator which converges at a substantially faster rate than in standard multivariate density estimation from i.i.d. observations.

Article information

Source
Ann. Statist., Volume 46, Number 6B (2018), 3451-3480.

Dates
Revised: October 2017
First available in Project Euclid: 11 September 2018

https://projecteuclid.org/euclid.aos/1536631280

Digital Object Identifier
doi:10.1214/17-AOS1664

Mathematical Reviews number (MathSciNet)
MR3852658

Zentralblatt MATH identifier
06965694

Citation

Strauch, Claudia. Adaptive invariant density estimation for ergodic diffusions over anisotropic classes. Ann. Statist. 46 (2018), no. 6B, 3451--3480. doi:10.1214/17-AOS1664. https://projecteuclid.org/euclid.aos/1536631280

References

• [1] Bakry, D., Bolley, F. and Gentil, I. (2011). Around Nash inequalities. In Journées Équations aux Dérivées Partielles (2010), Biarritz, France.
• [2] Bakry, D., Cattiaux, P. and Guillin, A. (2008). Rate of convergence for ergodic continuous Markov processes: Lyapunov versus Poincaré. J. Funct. Anal. 254 727–759.
• [3] Bakry, D., Gentil, I. and Ledoux, M. (2014). Analysis and Geometry of Markov Diffusion Operators. Grundlehren der Mathematischen Wissenschaften 348. Springer, Cham.
• [4] Carlen, E. A., Kusuoka, S. and Stroock, D. W. (1987). Upper bounds for symmetric Markov transition functions. Ann. Inst. Henri Poincaré Probab. Stat. 23 245–287.
• [5] Cattiaux, P. and Guillin, A. (2008). Deviation bounds for additive functionals of Markov processes. ESAIM Probab. Stat. 12 12–29.
• [6] Chorowski, J. and Trabs, M. (2016). Spectral estimation for diffusions with random sampling times. Stochastic Process. Appl. 126 2976–3008.
• [7] Dalalyan, A. and Reiß, M. (2007). Asymptotic statistical equivalence for ergodic diffusions: The multidimensional case. Probab. Theory Related Fields 137 25–47.
• [8] Davies, E. B. (1990). Heat Kernels and Spectral Theory. Cambridge Tracts in Mathematics 92. Cambridge Univ. Press, Cambridge.
• [9] Gobet, E., Hoffmann, M. and Reiß, M. (2004). Nonparametric estimation of scalar diffusions based on low frequency data. Ann. Statist. 32 2223–2253.
• [10] Goldenshluger, A. and Lepski, O. (2008). Universal pointwise selection rule in multivariate function estimation. Bernoulli 14 1150–1190.
• [11] Goldenshluger, A. and Lepski, O. (2009). Structural adaptation via $\mathbb{L}_{p}$-norm inequalities. Probab. Theory Related Fields 143 41–71.
• [12] Goldenshluger, A. and Lepski, O. (2011). Bandwidth selection in kernel density estimation: Oracle inequalities and adaptive minimax optimality. Ann. Statist. 39 1608–1632.
• [13] Kutoyants, Y. A. (2004). Statistical Inference for Ergodic Diffusion Processes. Springer, New York.
• [14] Ledoux, M. (2001). Logarithmic Sobolev inequalities for unbounded spin systems revisited. In Séminaire de Probabilités XXXV. Lecture Notes in Math. 1755 167–194. Springer, Berlin.
• [15] Lepski, O. (2013). Multivariate density estimation under sup-norm loss: Oracle approach, adaptation and independence structure. Ann. Statist. 41 1005–1034.
• [16] Lepski, O. (2013). Upper functions for positive random functionals. I. General setting and Gaussian random functions. Math. Methods Statist. 22 1–27.
• [17] Lepski, O. (2013). Upper functions for positive random functionals. II. Application to the empirical processes theory, Part 1. Math. Methods Statist. 22 83–99.
• [18] Lepski, O. (2013). Upper functions for positive random functionals. II. Application to the empirical processes theory, Part 2. Math. Methods Statist. 22 193–212.
• [19] Lepski, O. V. and Levit, B. Y. (1999). Adaptive nonparametric estimation of smooth multivariate functions. Math. Methods Statist. 8 344–370.
• [20] Lezaud, P. (2001). Chernoff and Berry–Esséen inequalities for Markov processes. ESAIM Probab. Stat. 5 183–201.
• [21] Nash, J. (1958). Continuity of solutions of parabolic and elliptic equations. Amer. J. Math. 80 931–954.
• [22] Nelson, E. (1958). The adjoint Markoff process. Duke Math. J. 25 671–690.
• [23] Nikolskii, S. M. (1975). Approximation of Functions of Several Variables and Imbedding Theorems. Die Grundlehren der Mathematischen Wissenschaften 205. Springer, New York. Translated from the Russian by J. M. Danskin, Jr.
• [24] Paulin, D. (2015). Concentration inequalities for Markov chains by Marton couplings and spectral methods. Electron. J. Probab. 20 1–32.
• [25] Söhl, J. and Trabs, M. (2016). Adaptive confidence bands for Markov chains and diffusions: Estimating the invariant measure and the drift. ESAIM Probab. Stat. 20 432–462.
• [26] Strauch, C. (2018). Supplement to “Adaptive invariant density estimation for ergodic diffusions over anisotropic classes.” DOI:10.1214/17-AOS1664SUPP.
• [27] Van der Vaart, A. W. and van Zanten, H. (2005). Donsker theorems for diffusions: Necessary and sufficient conditions. Ann. Probab. 33 1422–1451.

Supplemental materials

• Supplement to “Adaptive invariant density estimation for ergodic diffusions over anisotropic classes”. The supplementary file contains the proofs of the $\mathrm{sup}$-norm oracle-type inequality stated in Theorem 3.3, the upper bound in Theorem 3.4, the lower bound from Theorem 3.6 and of the results for estimators based on discrete observations stated in Theorem 4.1 and Theorem 4.3, respectively.