Pattern retrieval in an asymmetric neural network with embedded limit cycles
- 7 June 1989
- journal article
- Published by IOP Publishing in Journal of Physics A: General Physics
- Vol. 22 (11) , L525-L532
- https://doi.org/10.1088/0305-4470/22/11/013
Abstract
A neural network model in which the connection matrix is formed by summing direct products of successive patterns in cyclic sequences is studied. It is found that the noise elimination performance of this model can be better than that of an autocorrelation model due to a reduced tendency for trapping in spurious attractors.Keywords
This publication has 25 references indexed in Scilit:
- A Learning Algorithm for Boltzmann Machines*Published by Wiley ,2010
- Statistical mechanics of neural networks near saturationPublished by Elsevier ,2004
- Storage and retrieval of complex sequences in neural networksPhysical Review A, 1988
- Processing of Temporal Sequences in Neural NetworksPhysical Review Letters, 1988
- Layered feed-forward neural network with exactly soluble dynamicsPhysical Review A, 1988
- Noise-Driven Temporal Association in Neural NetworksEurophysics Letters, 1987
- Temporal Association in Asymmetric Neural NetworksPhysical Review Letters, 1986
- Collective computational properties of neural networks: New learning mechanismsPhysical Review A, 1986
- The Statistical Properties of the Hopfield Model of MemoryEurophysics Letters, 1986
- Learning and pattern recognition in spin glass modelsZeitschrift für Physik B Condensed Matter, 1985