Quality and efficiency of retrieval for Willshaw-like autoassociative networks: III. Willshaw–Potts model
A Kartashov, Alexander Frolov, Alexander Goltsev, R. Folk · Network Computation in Neural Systems · 1997
We study the recognition and correction properties of a structured associative memory with a floating threshold which can be considered a modification of the Willshaw network and bears some resemblance to a network of Potts neurons. An analytical study is performed for the asymptotic case of large networks and single-step retrieval. The main informational characteristics of the network are obtained; the number of spurious stable states and the size of their attraction basins are calculated. A comparison is made with the Willshaw network considered earlier in this series of papers.