Recurrent Associative Memory Network of Nonlinear Coupled Oscillators

Abstract. The recurrent associative memory networks with complexvalued Hebbian matrices of connections are designed from interacting limitcycle oscillators. These oscillatory networks have peculiarities and advantages as compared to Hop eld neural network model. In particular, the class of networks with high memory characteristics (the capacity close to 1, low extraneous memory) exists. At zero values of oscillator frequencies the designed networks are closely related to the known "clock" neural networks (networks from complex-valued neurons). Pattern recognition of colored images and recognition of objects with complicated topological structure look quite natural in the context of such models. Exact solutions have been obtained for a few types of the networks considered, in particular, for homogeneous closes chains.

• Type: Conference
• Year: 1997
• Where: ICANN
• Authors: Margarita Kuzmina, Eduard A. Manykin, Irina Surina
• Keywords: Associative Memory Networks, ICANN 1997, Neural Networks, Oscillatory Networks