English

Inference and learning in sparse systems with multiple states

Disordered Systems and Neural Networks 2013-09-03 v1

Abstract

We discuss how inference can be performed when data are sampled from the non-ergodic phase of systems with multiple attractors. We take as model system the finite connectivity Hopfield model in the memory phase and suggest a cavity method approach to reconstruct the couplings when the data are separately sampled from few attractor states. We also show how the inference results can be converted into a learning protocol for neural networks in which patterns are presented through weak external fields. The protocol is simple and fully local, and is able to store patterns with a finite overlap with the input patterns without ever reaching a spin glass phase where all memories are lost.

Keywords

Cite

@article{arxiv.1104.2775,
  title  = {Inference and learning in sparse systems with multiple states},
  author = {A. Braunstein and A. Ramezanpour and R. Zecchina and P. Zhang},
  journal= {arXiv preprint arXiv:1104.2775},
  year   = {2013}
}

Comments

15 pages, 10 figures, to be published in Phys. Rev. E

R2 v1 2026-06-21T17:54:05.792Z