Summary
2011 International Symposium on Nonlinear Theory and Its Applications
2011
Session Number:B1L-D
Session:
Number:B1L-D2
Associative Dynamics with Structured Relations Among Memories in a Chaotic Neural Network
Makito Oku, Kazuyuki Aihara,
pp.366-369
Publication Date:2011/9/4
Online ISSN:2188-5079
Summary:
The memories in our minds have associative relations among them, which form a complex network. Such a memory relation network (MRN) may be reflected in spontaneous transitions among brain states during resting and sleeping. Although such transitory dynamics can be described by the chaotic neural network (CNN) model, the influence of an MRN on the dynamics of the CNN model has not yet been investigated. Here, we consider a CNN with two types of connections. The first type is that of the auto-associative connections, which represent the memories themselves, whereas the second type is that of the hetero-associative connections, which represent the relations among these memories. We show numerically that the CNN partially follows a given MRN and chaotically visits the memories in some parameter regions. Our methods provide a way of introducing the MRN to the CNN model without destroying chaos.