Abstract
This paper proposes using a generalized Hopfield's model, also known as the McCulloch-Pitts model, as associative memories. The system always converges to the nearest stored words. This is not true for Hopfield's model. We also determine the information capacity of this generalized model. Our result is better than those in the literature.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 301-309 |
| Number of pages | 9 |
| Journal | Neural Networks |
| Volume | 3 |
| Issue number | 3 |
| DOIs | |
| State | Published - 1990 |
Keywords
- Associative memory
- Memory capacity
- Neural network model
ASJC Scopus subject areas
- Cognitive Neuroscience
- Artificial Intelligence
Fingerprint Dive into the research topics of 'Constructing associative memories using neural networks'. Together they form a unique fingerprint.
Cite this
Xu, X., & Tsai, W. T. (1990). Constructing associative memories using neural networks. Neural Networks, 3(3), 301-309. https://doi.org/10.1016/0893-6080(90)90073-T