Sequential Bayesian estimation for adaptive classification

Ji Won Yoon; Stephen J. Roberts; Matt Dyson; John Q. Gan

doi:10.1109/MFI.2008.4648010

Sequential Bayesian estimation for adaptive classification

Ji Won Yoon, Stephen J. Roberts, Matt Dyson, John Q. Gan

Research output: Contribution to conference › Paper › peer-review

9 Citations (Scopus)

Abstract

This paper proposes a robust algorithm to adapt a model for EEG signal classification using a modified Extended Kalman Filter (EKF). By applying Bayesian conjugate priors and marginalising the parameters, we can avoid the needs to estimate the covariances of the observation and hidden state noises. In addition, Laplace approximation is employed in our model to approximate non-Gaussian distributions as Gaussians.

Original language	English
Pages	601-605
Number of pages	5
DOIs	https://doi.org/10.1109/MFI.2008.4648010
Publication status	Published - 2008
Externally published	Yes
Event	2008 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, MFI - Seoul, Korea, Republic of Duration: 2008 Aug 20 → 2008 Aug 22

Other

Other	2008 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, MFI
Country/Territory	Korea, Republic of
City	Seoul
Period	08/8/20 → 08/8/22

Keywords

Extended Kalman filter
Laplace approximation
Marginalisation
Nonlinear dynamics

ASJC Scopus subject areas

Control and Systems Engineering
Software
Computer Science Applications

Access to Document

10.1109/MFI.2008.4648010

Cite this

@conference{6ce5ebd48a6e4fb2b9503236e20d8be3,

title = "Sequential Bayesian estimation for adaptive classification",

abstract = "This paper proposes a robust algorithm to adapt a model for EEG signal classification using a modified Extended Kalman Filter (EKF). By applying Bayesian conjugate priors and marginalising the parameters, we can avoid the needs to estimate the covariances of the observation and hidden state noises. In addition, Laplace approximation is employed in our model to approximate non-Gaussian distributions as Gaussians.",

keywords = "Extended Kalman filter, Laplace approximation, Marginalisation, Nonlinear dynamics",

author = "Yoon, {Ji Won} and Roberts, {Stephen J.} and Matt Dyson and Gan, {John Q.}",

year = "2008",

doi = "10.1109/MFI.2008.4648010",

language = "English",

pages = "601--605",

note = "2008 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, MFI ; Conference date: 20-08-2008 Through 22-08-2008",

}

TY - CONF

T1 - Sequential Bayesian estimation for adaptive classification

AU - Yoon, Ji Won

AU - Roberts, Stephen J.

AU - Dyson, Matt

AU - Gan, John Q.

PY - 2008

Y1 - 2008

N2 - This paper proposes a robust algorithm to adapt a model for EEG signal classification using a modified Extended Kalman Filter (EKF). By applying Bayesian conjugate priors and marginalising the parameters, we can avoid the needs to estimate the covariances of the observation and hidden state noises. In addition, Laplace approximation is employed in our model to approximate non-Gaussian distributions as Gaussians.

AB - This paper proposes a robust algorithm to adapt a model for EEG signal classification using a modified Extended Kalman Filter (EKF). By applying Bayesian conjugate priors and marginalising the parameters, we can avoid the needs to estimate the covariances of the observation and hidden state noises. In addition, Laplace approximation is employed in our model to approximate non-Gaussian distributions as Gaussians.

KW - Extended Kalman filter

KW - Laplace approximation

KW - Marginalisation

KW - Nonlinear dynamics

UR - http://www.scopus.com/inward/record.url?scp=67650530336&partnerID=8YFLogxK

U2 - 10.1109/MFI.2008.4648010

DO - 10.1109/MFI.2008.4648010

M3 - Paper

AN - SCOPUS:67650530336

SP - 601

EP - 605

T2 - 2008 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, MFI

Y2 - 20 August 2008 through 22 August 2008

ER -

Sequential Bayesian estimation for adaptive classification

Abstract

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this