Recovery of Performance in a Partially Connected Associative Memory Network Through Coding
Introducing partial connectivity to an associative memory network increases the variance of the dendritic sum distributions, reducing the performance. A coding scheme to compensate for this effect is considered, in which output patterns are self-organised by the network. It is shown using signal-to-noise ratio analysis that when the output patterns are self-organised the performance is greater than in a network with a higher connectivity and random patterns, in the regime of low connectivity and a high memory load. This analysis is supported by simulations. The self-organising network also outperforms the random network with input activity-dependent thresholding mechanisms in simulations.
Unable to display preview. Download preview PDF.
- Buckingham, J.: Delicate nets, faint recollections: a study of partially connected associative network memories. PhD thesis, Univ. Edinburgh (1991)Google Scholar