We discuss a neural network for the storage and retrieval of sequences of patterns. Assuming a distribution of delays, short relative to the duration of each of the single patterns, and assuming adaptive thresholds, we show that a sequence of patterns, learned at a particular speed, can be replayed
Sequences of smoothly correlated patterns in neural networks with random transmission delays
โ Scribed by M. Bartholomeus; A. C. C. Coolen
- Publisher
- Springer-Verlag
- Year
- 1992
- Tongue
- English
- Weight
- 909 KB
- Volume
- 67
- Category
- Article
- ISSN
- 0340-1200
No coin nor oath required. For personal study only.
โฆ Synopsis
A neural network with a broad distribution of transmission delays was used to study numerically the retrieval of sequences having several types of correlations between successive patterns. In the case of sequences consisting of patterns correlated for finite time, the quality of retrieval was found to be (more or less) independent of the pattern correlation width and the delay distribution. On the other hand the quality of retrieval is dependent on these factors in the case of sequences with pattern correlation functions with long time 'tails'. Finally we have studied to what extent the storage capacity depends on the pattern correlation function and the delay distribution.
๐ SIMILAR VOLUMES