Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing

Yoann Altmann, Alessandro Perelli, Mike E. Davies

Research output: Chapter in Book/Report/Conference proceedingConference contribution

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 per- formance based on EP approximations is compared to the per- formance using Monte Carlo sampling from the exact poste- rior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable pos- terior means and marginal variances.
LanguageEnglish
Title of host publication2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
PublisherIEEE
Pages5067-5071
Number of pages5
ISBN (Electronic)9781479981311
DOIs
Publication statusPublished - 17 Apr 2019
Event44th IEEE International Conference on Acoustics, Speech, and Signal Processing 2019 - Brighton, United Kingdom
Duration: 12 May 201917 May 2019

Conference

Conference44th IEEE International Conference on Acoustics, Speech, and Signal Processing 2019
Abbreviated titleICASSP 2019
CountryUnited Kingdom
CityBrighton
Period12/05/1917/05/19

Fingerprint

Linear regression
Photons
Sampling

Keywords

  • Approximate Bayesian inference
  • Expectation-Propagation
  • Poisson noise
  • linear regression

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Altmann, Y., Perelli, A., & Davies, M. E. (2019). Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) (pp. 5067-5071). [8682479] IEEE. https://doi.org/10.1109/ICASSP.2019.8682479
Altmann, Yoann ; Perelli, Alessandro ; Davies, Mike E. / Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing. 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). IEEE, 2019. pp. 5067-5071
@inproceedings{37522d0c952e4c149076f4925a1f8b99,
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 per- formance based on EP approximations is compared to the per- formance using Monte Carlo sampling from the exact poste- rior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable pos- terior means and marginal variances.",
keywords = "Approximate Bayesian inference, Expectation-Propagation, Poisson noise, linear regression",
author = "Yoann Altmann and Alessandro Perelli and Davies, {Mike E.}",
year = "2019",
month = "4",
day = "17",
doi = "10.1109/ICASSP.2019.8682479",
language = "English",
pages = "5067--5071",
booktitle = "2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)",
publisher = "IEEE",
address = "United States",

}

Altmann, Y, Perelli, A & Davies, ME 2019, Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing. in 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)., 8682479, IEEE, pp. 5067-5071, 44th IEEE International Conference on Acoustics, Speech, and Signal Processing 2019, Brighton, United Kingdom, 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.

2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). IEEE, 2019. p. 5067-5071 8682479.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing

AU - Altmann, Yoann

AU - Perelli, Alessandro

AU - Davies, Mike E.

PY - 2019/4/17

Y1 - 2019/4/17

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 per- formance based on EP approximations is compared to the per- formance using Monte Carlo sampling from the exact poste- rior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable pos- terior 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 per- formance based on EP approximations is compared to the per- formance using Monte Carlo sampling from the exact poste- rior distribution. In particular, we observe that using locally independent or isotropic approximate factors enables more robust and scalable algorithms while providing reliable pos- terior means and marginal variances.

KW - Approximate Bayesian inference

KW - Expectation-Propagation

KW - Poisson noise

KW - linear regression

UR - http://www.scopus.com/inward/record.url?scp=85068957633&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2019.8682479

DO - 10.1109/ICASSP.2019.8682479

M3 - Conference contribution

SP - 5067

EP - 5071

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

PB - IEEE

ER -

Altmann Y, Perelli A, Davies ME. Expectation-Propagation algorithms for linear regression with Poisson noise: application to photon-limited spectral unmixing. In 2019 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP). IEEE. 2019. p. 5067-5071. 8682479 https://doi.org/10.1109/ICASSP.2019.8682479