A non-linear explicit filter

被引：11

作者：

Genon-Catalot, V ^{[1
]}

机构：

[1] Univ Paris 05, UFR Math & Informat, F-75270 Paris 06, France

来源：

STATISTICS & PROBABILITY LETTERS | 2003年 / 61卷 / 02期

关键词：

filtering; discrete time observations; hidden Markov models; prior and posterior distributions;

D O I：

10.1016/S0167-7152(02)00344-9

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

We propose a non-linear model of filtering in discrete time. The signal is linear but noises are multiplicative and non-Gaussian. We exhibit a class of distributions including the standard centered Gaussian which is closed for the filtering algorithm. (C) 2003 Elsevier Science B.V. All rights reserved.

引用

页码：145 / 154

页数：10

共 9 条

[1]

[Anonymous], 1977, Statistics of Random Processes

[2]

Bickel P.J., 1996, BERNOULLI, V2, P199

[3]

DELMORAL P, 2000, IN PRESS LECT NOTES

[4] Asymptotics of the maximum likelihood estimator for general hidden Markov models [J].

Douc, R ;

Matias, C .

BERNOULLI, 2001, 7 (03) :381-420

[5]

Elliott R., 1995, Hidden Markov Models-Estimation and Control, V29

[6] Stochastic volatility models as hidden Markov models and statistical applications [J].

Genon-Catalot, V ;

Jeantheau, T ;

Larédo, C .

BERNOULLI, 2000, 6 (06) :1051-1079

[7]

GHYSELS E, 1996, HDB STAT, V14, P110

[8]

Jensen JL, 1999, ANN STAT, V27, P514

[9] MAXIMUM-LIKELIHOOD-ESTIMATION FOR HIDDEN MARKOV-MODELS [J].

LEROUX, BG .

STOCHASTIC PROCESSES AND THEIR APPLICATIONS, 1992, 40 (01) :127-143

← 1 →