Optimal storage of a neural network model: a replica symmetry-breaking solution
- 1 January 1993
- journal article
- Published by IOP Publishing in Journal of Physics A: General Physics
- Vol. 26 (2) , L61-L68
- https://doi.org/10.1088/0305-4470/26/2/007
Abstract
A break in replica symmetry is found above the critical line alpha c( kappa ) for storage with a finite minimal fraction of errors, within the first stage of the Parisi scheme, for noiseless networks with continuous synapses. The replica symmetry breaking solution yields the highest (most stable) minimal fraction of errors everywhere above alpha c( kappa ).Keywords
This publication has 10 references indexed in Scilit:
- Gardner-Derrida neural networks with correlated patternsJournal of Physics A: General Physics, 1991
- Learning and retrieval in attractor neural networks above saturationJournal of Physics A: General Physics, 1991
- Retrieval phase diagrams for attractor neural networks with optimal interactionsJournal of Physics A: General Physics, 1990
- Storage capacity of memory networks with binary couplingsJournal de Physique, 1989
- Optimal storage properties of neural network modelsJournal of Physics A: General Physics, 1988
- The space of interactions in neural network modelsJournal of Physics A: General Physics, 1988
- Spin glasses: Experimental facts, theoretical concepts, and open questionsReviews of Modern Physics, 1986
- Stability of a crystal-field split spin glassJournal of Physics C: Solid State Physics, 1985
- Field and anisotropy effects in spin glasses. I. The collinear caseJournal of Physics C: Solid State Physics, 1983
- The order parameter for spin glasses: a function on the interval 0-1Journal of Physics A: General Physics, 1980