Phase-space Gardening in the Binary-couplings Memory Network

Abstract

The dynamics of the binary-couplings Hoppeld/Gardner associative memory network is studied. A iterative learning rule is presented that allows to adjust the stabilities for each pattern at each neuron individually. Simulations of the resulting network show that the basins of attraction of the patterns can be shaped as desired. The dependence of m c on the stability is shown.

Cite this paper

@inproceedings{Hendrich1993PhasespaceGI, title={Phase-space Gardening in the Binary-couplings Memory Network}, author={Norman Hendrich}, year={1993} }