Work

Mean Field Theory for Sigmoid Belief Networks

Year: 1996

Type: article

Abstract: We develop a mean field theory for sigmoid belief networks based on ideas from statistical mechanics. Our mean field theory provides a tractable approximation to the true probability distribution in these networks; it also yields a lower bound on the likelihood of evidence. We demonstrate the utility of this framework on a benchmark problem in statistical pattern recognition---the classification of handwritten digits.

Source: Journal of Artificial Intelligence Research

DOI: 10.1613/jair.251

Authors: L. K. Saul, T. Jaakkola, M. I. Jordan

Institution: Massachusetts Institute of Technology

Language: English

Topic: Neural Networks and Applications

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

Funders: National Science Foundation, Siemens USA, University of Toronto, Office of Naval Research

Awards: Post Doctoral: Probabilistic Models for Hierarchical Neural Networks, G8876996369