Catalogue of Artificial Intelligence Techniques
Keywords: neural networks
Categories: Neural Networks
Author(s): Martin Westhead
Associative memories work by recalling information in response to an information cue, rather than an address (as in standard computer memory). Associative memories can be autoassociative--recalling the same information that is used as a cue, or heteroassociative--where the association is made with different information. The former can be used to complete a partial pattern, or clean up noisy data, the latter is more obviously useful as a memory. They can be implemented in many ways. For example Hopfield Nets can be used for autoassociative memory, and Willshaw Nets will do either.
- Willshaw, D.J., Buneman, O.P. and Longuet-Higgins, H.C., Non-Holographic Associative Memory (Anderson, J.A. and Rosenfeld, E.
, eds.), MIT Press, Cambridge, MA, 1988.