Categorization by a three-state attractor neural network

Abstract

The categorization properties of an attractor network of threestate neurons which infers three-state concepts from examples are studied. The evolution equations governing the parallel dynamics at zero temperature for the overlap between the state of the network and the examples, the state of the network and the concepts as well as the neuron activity are discussed in the limit of extreme dilution. A transition from a retrieval region to a categorization region is found when the number of examples or their correlations are increased. If the pattern activity is small enough, the examples (concepts) are very well retrieved (categorized) for an appropriate choice of the zero-activity threshold of the neurons.

Categorization in fully connected multistate neural network models

“…In the first case we also checked the robustness of the network performance to synaptic noise. Our results are restricted to binary ancestors and multi-state descendents, although the case of multi-state ancestors has been considered in an extremely dilute network [21].…”

“…More recently, some work has been done on the categorization problem in multi-state attractor networks [17][18][19][20][21], following extensive studies of the problem in binary networks [22][23][24][25][26][27][28][29][30][31]. The categorization problem consists in the spontaneous recognition of a level of hierarchical patterns other than those stored in the training process of a network [22,23].…”