Jump to content

Product of experts

From Wikipedia, the free encyclopedia

Product of experts (PoE) is a machine learning technique. It models a probability distribution by combining the output from several simpler distributions. It was proposed by Geoffrey Hinton in 1999,[1] along with an algorithm for training the parameters of such a system.

The core idea is to combine several probability distributions ("experts") by multiplying their density functions—making the PoE classification similar to an "and" operation. This allows each expert to make decisions on the basis of a few dimensions without having to cover the full dimensionality of a problem:

where are unnormalized expert densities and is a normalization constant (see partition function (statistical mechanics)).

This is related to (but quite different from) a mixture model, where several probability distributions are combined via an "or" operation, which is a weighted sum of their density functions: with

The experts may be understood as each being responsible for enforcing a constraint in a high-dimensional space. A data point is considered likely iff none of the experts say that the point violates a constraint.

To optimize it, he proposed the contrastive divergence minimization algorithm.[2] This algorithm is most often used for learning restricted Boltzmann machines.

See also

[edit]

References

[edit]
  1. ^ Hinton, G.E. (1999). "Products of experts". 9th International Conference on Artificial Neural Networks: ICANN '99. Vol. 1999. IEE. pp. 1–6. doi:10.1049/cp:19991075. ISBN 978-0-85296-721-8.
  2. ^ Hinton, Geoffrey E. (2002-08-01). "Training Products of Experts by Minimizing Contrastive Divergence". Neural Computation. 14 (8): 1771–1800. doi:10.1162/089976602760128018. ISSN 0899-7667. PMID 12180402. S2CID 207596505.
[edit]