Approximate inference in Boltzmann machines

Journal article

Inference in Boltzmann machines is NP-hard in general. As a result approximations are often necessary. We discuss first order mean field and second order Onsager truncations of the Plefka expansion of the Gibbs free energy. The Bethe free energy is introduced and rewritten as a Gibbs free energy. From there a convergent belief optimization algorithm is derived to minimize the Bethe free energy. An analytic expression for the linear response estimate of the covariances is found which is exact on Boltzmann trees. Finally, a number of theorems is proven concerning the Plefka expansion, relating the first order mean field and the second order Onsager approximation to the Bethe approximation. Experiments compare mean field approximation, Onsager approximation, belief propagation and belief optimization. © 2002 Elsevier Science B.V. All rights reserved.

Authors: Welling, M; Teh, Y

• Publication status: Published
• Journal: ARTIFICIAL INTELLIGENCE
• Volume: 143
• Issue: 1
• Pages: 19-50
• Publication date: 2003-01-01
• DOI: 10.1016/S0004-3702(02)00361-2
• ISSN: 0004-3702
• Language: English
• Keywords: Boltzmann machines; inference; advanced mean field methods; loopy belief propagation