Title | ||
---|---|---|
Generalized Hopfield networks for associative memories with multi-valued stable states |
Abstract | ||
---|---|---|
Hopfield networks with fully connected standard neurons can be generalized by replacing bi-level activation functions with their multilevel counterparts. Multilevel neuron characteristics are discussed in the paper with emphasis on their inflection points. It is shown that an activation function possessing (N + 1)-levels yields N + 1 minima and N saddle points of the computational energy function when two generalized neurons are used in a conventional bi-stable connection. Analytical results for parameter constraints and energy function properties are discussed for binary and ternary characteristics of neurons. Gradient fields indicating basins of attraction for continuous-time networks are used to illustrate dynamical relationships during network convergence to stable points. Results indicate that generalized Hopfield networks can be used for multilevel signal processing and smoothing of planar images. |
Year | DOI | Venue |
---|---|---|
1996 | 10.1016/0925-2312(96)00086-0 | Neurocomputing |
Keywords | Field | DocType |
Hopfield network,Associative memory,Multivalued neural networks,Multistable flip flops | Inflection point,Content-addressable memory,Saddle point,Activation function,Maxima and minima,Smoothing,Artificial intelligence,Hopfield network,Machine learning,Mathematics,Binary number | Journal |
Volume | Issue | ISSN |
13 | 2-4 | 0925-2312 |
Citations | PageRank | References |
24 | 4.01 | 2 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jacek M. Zurada | 1 | 2553 | 226.22 |
Ian Cloete | 2 | 132 | 16.61 |
Etienne van der Poel | 3 | 24 | 4.01 |