TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models

Tutkimustuotosvertaisarvioitu

Standard

An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models. / Ala-Luhtala, Juha; Whiteley, Nick; Heine, Kari; Piche, Robert.

julkaisussa: IEEE Transactions on Signal Processing, Vuosikerta 64, Nro 18, 05.05.2016, s. 4875-4890.

Tutkimustuotosvertaisarvioitu

Harvard

Ala-Luhtala, J, Whiteley, N, Heine, K & Piche, R 2016, 'An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models' IEEE Transactions on Signal Processing, Vuosikerta. 64, Nro 18, Sivut 4875-4890. https://doi.org/10.1109/TSP.2016.2563387

APA

Ala-Luhtala, J., Whiteley, N., Heine, K., & Piche, R. (2016). An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models. IEEE Transactions on Signal Processing, 64(18), 4875-4890. https://doi.org/10.1109/TSP.2016.2563387

Vancouver

Ala-Luhtala J, Whiteley N, Heine K, Piche R. An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models. IEEE Transactions on Signal Processing. 2016 touko 5;64(18):4875-4890. https://doi.org/10.1109/TSP.2016.2563387

Author

Ala-Luhtala, Juha ; Whiteley, Nick ; Heine, Kari ; Piche, Robert. / An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models. Julkaisussa: IEEE Transactions on Signal Processing. 2016 ; Vuosikerta 64, Nro 18. Sivut 4875-4890.

Bibtex - Lataa

@article{c5eac1a553b14ada806eb591079da7be,
title = "An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models",
abstract = "Twisted particle filters are a class of sequential Monte Carlo methods recently introduced by Whiteley and Lee to improve the efficiency of marginal likelihood estimation in state-space models. The purpose of this article is to extend the twisted particle filtering methodology, establish accessible theoretical results which convey its rationale, and provide a demonstration of its practical performance within particle Markov chain Monte Carlo for estimating static model parameters. We derive twisted particle filters that incorporate systematic or multinomial resampling and information from historical particle states, and a transparent proof which identifies the optimal algorithm for marginal likelihood estimation. We demonstrate how to approximate the optimal algorithm for nonlinear state-space models with Gaussian noise and we apply such approximations to two examples: a range and bearing tracking problem and an indoor positioning problem with Bluetooth signal strength measurements. We demonstrate improvements over standard algorithms in terms of variance of marginal likelihood estimates and Markov chain autocorrelation for given CPU time, and improved tracking performance using estimated parameters.",
author = "Juha Ala-Luhtala and Nick Whiteley and Kari Heine and Robert Piche",
note = "EXT={"}Ala-Luhtala, Juha{"}",
year = "2016",
month = "5",
day = "5",
doi = "10.1109/TSP.2016.2563387",
language = "English",
volume = "64",
pages = "4875--4890",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC",
number = "18",

}

RIS (suitable for import to EndNote) - Lataa

TY - JOUR

T1 - An Introduction to Twisted Particle Filters and Parameter Estimation in Non-Linear State-Space Models

AU - Ala-Luhtala, Juha

AU - Whiteley, Nick

AU - Heine, Kari

AU - Piche, Robert

N1 - EXT="Ala-Luhtala, Juha"

PY - 2016/5/5

Y1 - 2016/5/5

N2 - Twisted particle filters are a class of sequential Monte Carlo methods recently introduced by Whiteley and Lee to improve the efficiency of marginal likelihood estimation in state-space models. The purpose of this article is to extend the twisted particle filtering methodology, establish accessible theoretical results which convey its rationale, and provide a demonstration of its practical performance within particle Markov chain Monte Carlo for estimating static model parameters. We derive twisted particle filters that incorporate systematic or multinomial resampling and information from historical particle states, and a transparent proof which identifies the optimal algorithm for marginal likelihood estimation. We demonstrate how to approximate the optimal algorithm for nonlinear state-space models with Gaussian noise and we apply such approximations to two examples: a range and bearing tracking problem and an indoor positioning problem with Bluetooth signal strength measurements. We demonstrate improvements over standard algorithms in terms of variance of marginal likelihood estimates and Markov chain autocorrelation for given CPU time, and improved tracking performance using estimated parameters.

AB - Twisted particle filters are a class of sequential Monte Carlo methods recently introduced by Whiteley and Lee to improve the efficiency of marginal likelihood estimation in state-space models. The purpose of this article is to extend the twisted particle filtering methodology, establish accessible theoretical results which convey its rationale, and provide a demonstration of its practical performance within particle Markov chain Monte Carlo for estimating static model parameters. We derive twisted particle filters that incorporate systematic or multinomial resampling and information from historical particle states, and a transparent proof which identifies the optimal algorithm for marginal likelihood estimation. We demonstrate how to approximate the optimal algorithm for nonlinear state-space models with Gaussian noise and we apply such approximations to two examples: a range and bearing tracking problem and an indoor positioning problem with Bluetooth signal strength measurements. We demonstrate improvements over standard algorithms in terms of variance of marginal likelihood estimates and Markov chain autocorrelation for given CPU time, and improved tracking performance using estimated parameters.

U2 - 10.1109/TSP.2016.2563387

DO - 10.1109/TSP.2016.2563387

M3 - Article

VL - 64

SP - 4875

EP - 4890

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 18

ER -