General potential surfaces and neural networks
- 1 March 1988
- journal article
- research article
- Published by American Physical Society (APS) in Physical Review A
- Vol. 37 (6), 2134-2143
- https://doi.org/10.1103/physreva.37.2134
Abstract
Investigation of Hopfield’s model of associative-memory implementation by a neural network led to an associative-memory model based on a generalized potential surface. In this model, there are no spurious memories, and any set of desired points can be stored with unlimited capacity (in the continuous-time real-space version of the model). There are no limit cycles in this system, and the size of all basins of attraction can reach up to half the distance between stored points by proper choice of the design parameters. A discrete-time version with its state-space being the unit hypercube is also derived, and admits a worst-case capacity (under any fixed desired size of basins of attractions) which grows exponentially with the number of neurons at a rate that is asymptotically optimal in the information theory sense. The computational complexity of this model is similar to that of the Hopfield memory. The results are derived under an axiomatic approach which determines the desired properties and shows that the above-mentioned model is the only one to achieve them.Keywords
This publication has 12 references indexed in Scilit:
- The capacity of the Hopfield associative memoryIEEE Transactions on Information Theory, 1987
- Number of stable points for spin-glasses and neural networks of higher ordersPhysical Review Letters, 1987
- Associative recall of memory without errorsPhysical Review A, 1987
- Temporal Association in Asymmetric Neural NetworksPhysical Review Letters, 1986
- Sequential state generation by model neural networks.Proceedings of the National Academy of Sciences, 1986
- Structure of metastable states in the Hopfield modelJournal of Physics A: General Physics, 1986
- Storing Infinite Numbers of Patterns in a Spin-Glass Model of Neural NetworksPhysical Review Letters, 1985
- Spin-glass models of neural networksPhysical Review A, 1985
- Information capacity of the Hopfield modelIEEE Transactions on Information Theory, 1985
- Neural networks and physical systems with emergent collective computational abilities.Proceedings of the National Academy of Sciences, 1982