Optimal capacity of graded-response perceptrons: a replica-symmetry-breaking solution
- 21 May 1996
- journal article
- Published by IOP Publishing in Journal of Physics A: General Physics
- Vol. 29 (10) , 2299-2307
- https://doi.org/10.1088/0305-4470/29/10/010
Abstract
Optimal capacities of perceptrons with graded input - output relations are studied within the first-step replica-symmetry-breaking Gardner approach. Input-data errors and a limited output precision are allowed. In particular, the role of non-monotonicity in the input - output relations on the breaking and on the overall performance is determined.Keywords
This publication has 19 references indexed in Scilit:
- The Tenet real-time protocol suite: design, implementation, and experiencesIEEE/ACM Transactions on Networking, 1996
- On the dynamics of analogue neurons with nonsigmoidal gain functionsPhysica A: Statistical Mechanics and its Applications, 1996
- Response functions improving performance in analog attractor neural networksPhysical Review E, 1994
- Optimal signalling in attractor neural networksNetwork: Computation in Neural Systems, 1994
- Symmetry breaking in nonmonotonic neural networksJournal of Physics A: General Physics, 1993
- MEMORY RETRIEVAL IN OPTIMAL SUBSPACESInternational Journal of Neural Systems, 1992
- A real-time neural system for color constancyIEEE Transactions on Neural Networks, 1991
- 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
- Maximum Storage Capacity in Neural NetworksEurophysics Letters, 1987