Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing

Yoann Altmann; Alessandro Perelli; Mike E. Davies

doi:10.1109/ICASSP.2019.8682479

Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing

Yoann Altmann, Alessandro Perelli, Mike E. Davies

Biomedical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

7 Citations (Scopus)

Abstract

This paper discusses Expectation-Propagation (EP) methods for approximate Bayesian inference in the context of linear regression with Poisson noise. We review two main factor graphs used for generalized linear models and discuss how different EP algorithms can be derived. The estimation performance based on EP approximations is compared to the performance using Monte Carlo sampling from the exact posterior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable posterior means and marginal variances.

Original language	English
Title of host publication	ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Publisher	IEEE
Pages	5067-5071
Number of pages	5
ISBN (Print)	978-1-4799-8132-8
DOIs	https://doi.org/10.1109/ICASSP.2019.8682479
Publication status	Published - 17 May 2019
Event	ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) - Brighton, UK Duration: 12 May 2019 → 17 May 2019

Conference

Conference	ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Period	12/05/19 → 17/05/19

Keywords

Estimation
Covariance matrices
Bayes methods
Approximation algorithms
Signal processing algorithms
Gaussian distribution
Linear regression

Access to Document

10.1109/ICASSP.2019.8682479

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8682479

Cite this

@inproceedings{8f6a920b81c24b7ca3fa4ae3cb299efb,

title = "Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing",

abstract = "This paper discusses Expectation-Propagation (EP) methods for approximate Bayesian inference in the context of linear regression with Poisson noise. We review two main factor graphs used for generalized linear models and discuss how different EP algorithms can be derived. The estimation performance based on EP approximations is compared to the performance using Monte Carlo sampling from the exact posterior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable posterior means and marginal variances.",

keywords = "Estimation, Covariance matrices, Bayes methods, Approximation algorithms, Signal processing algorithms, Gaussian distribution, Linear regression",

author = "Yoann Altmann and Alessandro Perelli and Davies, {Mike E.}",

year = "2019",

month = may,

day = "17",

doi = "10.1109/ICASSP.2019.8682479",

language = "English",

isbn = "978-1-4799-8132-8",

pages = "5067--5071",

booktitle = "ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",

publisher = "IEEE",

note = "ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) ; Conference date: 12-05-2019 Through 17-05-2019",

}

Altmann, Y, Perelli, A & Davies, ME 2019, Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing. in ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)., 8682479, IEEE, pp. 5067-5071, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 12/05/19. https://doi.org/10.1109/ICASSP.2019.8682479

Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing. / Altmann, Yoann; Perelli, Alessandro; Davies, Mike E.
ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2019. p. 5067-5071 8682479.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Expectation-propagation algorithms for linear regression with poisson noise

T2 - ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

AU - Altmann, Yoann

AU - Perelli, Alessandro

AU - Davies, Mike E.

PY - 2019/5/17

Y1 - 2019/5/17

N2 - This paper discusses Expectation-Propagation (EP) methods for approximate Bayesian inference in the context of linear regression with Poisson noise. We review two main factor graphs used for generalized linear models and discuss how different EP algorithms can be derived. The estimation performance based on EP approximations is compared to the performance using Monte Carlo sampling from the exact posterior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable posterior means and marginal variances.

AB - This paper discusses Expectation-Propagation (EP) methods for approximate Bayesian inference in the context of linear regression with Poisson noise. We review two main factor graphs used for generalized linear models and discuss how different EP algorithms can be derived. The estimation performance based on EP approximations is compared to the performance using Monte Carlo sampling from the exact posterior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable posterior means and marginal variances.

KW - Estimation

KW - Covariance matrices

KW - Bayes methods

KW - Approximation algorithms

KW - Signal processing algorithms

KW - Gaussian distribution

KW - Linear regression

UR - https://ieeexplore.ieee.org/document/8682479/

U2 - 10.1109/ICASSP.2019.8682479

DO - 10.1109/ICASSP.2019.8682479

M3 - Conference contribution

SN - 978-1-4799-8132-8

SP - 5067

EP - 5071

BT - ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

PB - IEEE

Y2 - 12 May 2019 through 17 May 2019

ER -

Expectation-propagation algorithms for linear regression with poisson noise: Application to photon-limited spectral unmixing

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this