A Variational Bayesian Committee of Neural Networks

Neil D. Lawrence; Mehdi Azzouzi

edit

Back to publications

A Variational Bayesian Committee of Neural Networks

Neil D. Lawrence, Mehdi Azzouzi

, 1999.

Abstract

Exact inference in Bayesian neural networks is non analytic to compute and as a result approximate approaches such as the evidence procedure, Monte-Carlo sampling and variational inference have been proposed. In this paper we present a general overview of the Bayesian approach with a particular emphasis on the variational procedure. We then present a new approximating distribution based on mixtures of Gaussian distributions and show how it may be implemented. We present results on a simple toy problem and on two real world data-sets.

Links

Cite this Paper

BibTeX


@Misc{Lawrence:nnmixtures99,
  title = 	 {A Variational Bayesian Committee of Neural Networks},
  author = 	 {Lawrence, Neil D. and Azzouzi, Mehdi},
  year = 	 {1999},
  pdf = 	 {http://www.thelawrences.net/neil/nnmixture.pdf},
  url = 	 {/publications/lawrence-nnmixtures99.html},
  abstract = 	 {Exact inference in Bayesian neural networks is non analytic to compute and
as a result approximate approaches such as the evidence procedure, Monte-Carlo sampling
and variational inference have been proposed. In this paper we present a general
overview of the Bayesian approach with a particular emphasis on the variational
procedure. We then present a new approximating distribution based on *mixtures*
of Gaussian distributions and show how it may be implemented. We present results
on a simple toy problem and on two real world data-sets.
}
}

Endnote

%0 Generic
%T A Variational Bayesian Committee of Neural Networks
%A Neil D. Lawrence
%A Mehdi Azzouzi
%D 1999	
%F Lawrence:nnmixtures99
%U /publications/lawrence-nnmixtures99.html
%X Exact inference in Bayesian neural networks is non analytic to compute and
as a result approximate approaches such as the evidence procedure, Monte-Carlo sampling
and variational inference have been proposed. In this paper we present a general
overview of the Bayesian approach with a particular emphasis on the variational
procedure. We then present a new approximating distribution based on *mixtures*
of Gaussian distributions and show how it may be implemented. We present results
on a simple toy problem and on two real world data-sets.

RIS


TY  - GEN
TI  - A Variational Bayesian Committee of Neural Networks
AU  - Neil D. Lawrence
AU  - Mehdi Azzouzi
DA  - 1999/09/17	
ID  - Lawrence:nnmixtures99
L1  - http://www.thelawrences.net/neil/nnmixture.pdf
UR  - /publications/lawrence-nnmixtures99.html
AB  - Exact inference in Bayesian neural networks is non analytic to compute and
as a result approximate approaches such as the evidence procedure, Monte-Carlo sampling
and variational inference have been proposed. In this paper we present a general
overview of the Bayesian approach with a particular emphasis on the variational
procedure. We then present a new approximating distribution based on *mixtures*
of Gaussian distributions and show how it may be implemented. We present results
on a simple toy problem and on two real world data-sets.

ER  -

APA


Lawrence, N.D. & Azzouzi, M.. (1999). A Variational Bayesian Committee of Neural Networks.  Available from /publications/lawrence-nnmixtures99.html.