In this paper, we estimate the expected tracking error of a fixed gain stochastic approximation scheme. The underlying process is not assumed Markovian, a mixing condition is required instead. Furthermore, the updating function may be discontinuous in the parameter.
Accepté le :
DOI : 10.1051/ps/2018019
Mots-clés : Adaptive control, stochastic gradient, fixed gain, recursive estimators, parameter discontinuity, mixing processes, non-Markovian dynamics
@article{PS_2019__23__217_0,
author = {Chau, Huy N. and Kumar, Chaman and R\'asonyi, Mikl\'os and Sabanis, Sotirios},
title = {On fixed gain recursive estimators with discontinuity in the parameters},
journal = {ESAIM: Probability and Statistics},
pages = {217--244},
publisher = {EDP-Sciences},
volume = {23},
year = {2019},
doi = {10.1051/ps/2018019},
mrnumber = {3945579},
zbl = {1420.62359},
language = {en},
url = {http://www.numdam.org/articles/10.1051/ps/2018019/}
}
TY - JOUR AU - Chau, Huy N. AU - Kumar, Chaman AU - Rásonyi, Miklós AU - Sabanis, Sotirios TI - On fixed gain recursive estimators with discontinuity in the parameters JO - ESAIM: Probability and Statistics PY - 2019 SP - 217 EP - 244 VL - 23 PB - EDP-Sciences UR - http://www.numdam.org/articles/10.1051/ps/2018019/ DO - 10.1051/ps/2018019 LA - en ID - PS_2019__23__217_0 ER -
%0 Journal Article %A Chau, Huy N. %A Kumar, Chaman %A Rásonyi, Miklós %A Sabanis, Sotirios %T On fixed gain recursive estimators with discontinuity in the parameters %J ESAIM: Probability and Statistics %D 2019 %P 217-244 %V 23 %I EDP-Sciences %U http://www.numdam.org/articles/10.1051/ps/2018019/ %R 10.1051/ps/2018019 %G en %F PS_2019__23__217_0
Chau, Huy N.; Kumar, Chaman; Rásonyi, Miklós; Sabanis, Sotirios. On fixed gain recursive estimators with discontinuity in the parameters. ESAIM: Probability and Statistics, Tome 23 (2019), pp. 217-244. doi : 10.1051/ps/2018019. http://www.numdam.org/articles/10.1051/ps/2018019/
[1] , and , Convergence of the one-dimensional Kohonen algorithm. Adv. Appl. Probab. 30 (1998) 850–869. | DOI | MR | Zbl
[2] and , An approach to the existence of unique invariant probabilities for Markov processes. János Bolyai Math. Soc. I (2002) 181–200. | MR | Zbl
[3] , and , Weak convergence and local stability properties of fixed step size recursive algorithms. IEEE Trans. Inf. Theory 39 (1993) 966–978. | DOI | MR | Zbl
[4] and , Long memory in continuous-time stochastic volatility models. Math. Financ. 8 (1998) 291–323. | DOI | MR | Zbl
[5] and , Iterated random functions. SIAM Rev. 41 (1999) 45–76. | DOI | MR | Zbl
[6] and , Markov Processes: Characterization and Convergence. Wiley, New York (1986). | DOI | MR | Zbl
[7] , Convergence analysis of an adaptive filter equipped with the sign-sign algorithm. IEEE Trans. Automat. Control 40 (1995) 1807–1811. | DOI | MR | Zbl
[8] , Analysis and design of a signed regressor LMS algorithm for stationary and nonstationary adaptive filtering with correlated Gaussian data. IEEE Trans. Circ. Syst. 37 (1990) 1367–1374. | DOI
[9] , , and , Convergence of Markovian stochastic approximation with discontinuous dynamics. SIAM J. Control Optim. 54 (2016) 866–893. | DOI | MR | Zbl
[10] , and , Volatility is rough. Quant. Financ. 18 (2018) 933–949. | DOI | MR | Zbl
[11] , Some averaging and stability results for random differential equations. SIAM J. Appl. Math. 36 (1979) 86–105. | DOI | MR | Zbl
[12] , On a class of mixing processes. Stochastics 26 (1989) 165–191. | MR | Zbl
[13] , On fixed gain recursive estimation processes. Retrieval code for full electronic manuscript: 56854. J. Math. Syst. Estim. Control 6 (1996) 355–358. | MR | Zbl
[14] , Strong approximation of the recursive prediction error estimator of the parameters of an ARMA process. Syst. Control Lett. 21 (1993) 347–351. | DOI | MR | Zbl
[15] , On Rissanen’s predictive stochastic complexity for stationary ARMA processes. J. Stat. Plan. Inference 41 (1994) 303–325. | DOI | MR | Zbl
[16] , AR(∞) estimation and nonparametric stochastic complexity. IEEE Trans. Inf. Theory 38 (1992) 1768–1778. | DOI | MR | Zbl
[17] , Stability of random iterative mappings, in Modeling Uncertainty. An Examination of Stochastic Theory, Methods,and Applications, edited by , and . Vol. 46 of International Series in Operations Research and Management Science. Kluwer Academic Publishers, Boston (2002) 359–371. | MR
[18] , A representation theorem for the error of recursive estimators. SIAM J. Control Optim. 44 (2006) 2123–2188. | DOI | MR | Zbl
[19] , Rate of convergence of recursive estimators. SIAM J. Control Optim. 30 (1992) 1200–1227. | DOI | MR | Zbl
[20] , , , and , New methods for the statistical analysis of Hidden Markov models, in Proceedings of the 41st IEEE Conference on Decision and Control (2002), Las Vegas, USA. IEEE Press, New York (2002) 2272–2277. | DOI
[21] , and , Large Sample Inference for Long Memory Processes. Imperial College Press, London (2012). | DOI | MR | Zbl
[22] and , Yet another look at Harris’ ergodic theorem for Markov chains, in Seminar on Stochastic Analysis, Random Fields and Applications VI, edited by , and . Vol. 63 of Progress in Probability. Springer, Basel (2011) 109–117. | MR | Zbl
[23] , Ordinary differential equations. Vol. 38 of Classics in Applied Mathematics, 2nd edition. SIAM, PA (2002). | MR | Zbl
[24] , Foundations of Modern Probability, 2nd edition. Springer, NY (2002). | DOI | MR | Zbl
[25] , Fourier Analysis in Probability Theory. Academic Press, NY (1972). | MR | Zbl
[26] and Regression quantiles. Econometrica 46 (1978) 33–50. | DOI | MR | Zbl
[27] , Analysis of a simple self-organising process. Biol. Cybern. 44 (1982) 135–140. | DOI | MR | Zbl
[28] and , Stochastic approximation with averaging innovation applied to finance. Monte Carlo Methods Appl. 18 (2012) 1–51. | DOI | MR | Zbl
[29] , Discrete-Parameter Martingales. North-Holland Mathematical Library. American Elsevier, NY (1975). | MR | Zbl
[30] , On the statistical analysis of quantized Gaussian AR(1) processes. Int. J. Adapt. Control Signal Process. 24 (2010) 490–507. | DOI | MR | Zbl
Cité par Sources :
