Optimal Belief Approximation

10/27/2016
by   Reimar H. Leike, et al.
0

In Bayesian statistics probability distributions express beliefs. However, for many problems the beliefs cannot be computed analytically and approximations of beliefs are needed. We seek a loss function that quantifies how "embarrassing" it is to communicate a given approximation. We reproduce and discuss an old proof showing that there is only one ranking under the requirements that (1) the best ranked approximation is the non-approximated belief and (2) that the ranking judges approximations only by their predictions for actual outcomes. The loss function that is obtained in the derivation is equal to the Kullback-Leibler divergence when normalized. This loss function is frequently used in the literature. However, there seems to be confusion about the correct order in which its functional arguments, the approximated and non-approximated beliefs, should be used. The correct order ensures that the recipient of a communication is only deprived of the minimal amount of information. We hope that the elementary derivation settles the apparent confusion. For example when approximating beliefs with Gaussian distributions the optimal approximation is given by moment matching. This is in contrast to many suggested computational schemes.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
02/27/2013

Ignorance and the Expressiveness of Single- and Set-Valued Probability Models of Belief

Over time, there have hen refinements in the way that probability distri...
research
06/02/2011

Learning unbelievable marginal probabilities

Loopy belief propagation performs approximate inference on graphical mod...
research
10/19/2012

Loopy Belief Propagation as a Basis for Communication in Sensor Networks

Sensor networks are an exciting new kind of computer system. Consisting ...
research
03/27/2013

Belief in Belief Functions: An Examination of Shafer's Canonical Examples

In the canonical examples underlying Shafer-Dempster theory, beliefs ove...
research
03/06/2013

Using Tree-Decomposable Structures to Approximate Belief Networks

Tree structures have been shown to provide an efficient framework for pr...
research
03/13/2013

Intuitions about Ordered Beliefs Leading to Probabilistic Models

The general use of subjective probabilities to model belief has been jus...
research
02/11/2022

Dandelion: Certified Approximations of Elementary Functions

Elementary function operations such as sin and exp cannot in general be ...

Please sign up or login with your details

Forgot password? Click here to reset