Continuous Attractors of 3-D Discrete-Time Ring Networks with Circulant Weight Matrix

Jiali Yu, Zhang Yi, Yong Liao, De-An Wu, Xiong Dai

Abstract

A continuous attractor of a recurrent neural network is a set of connected stable equilibrium points. The continuous attractors of neural networks with symmetric connection weights have been studied widely. However, most of the matrixes are asymmetric. In this paper, the continuous attractors of 3-D discrete-time ring networks with circulant weight matrix are studied. The circulant matrix is asymmetric. The eigenvalues of asymmetric matrix may be complex. Based on the complex eigenvalues, the conditions that guarantee the networks with circulant weight matrix to have continuous attractors are obtained.
BibTeX
@inproceedings{yu2018continuous,
  author = {Jiali Yu and Zhang Yi and Yong Liao and De-An Wu and Xiong Dai},
  title = {{Continuous Attractors of 3-D Discrete-Time Ring Networks with Circulant Weight Matrix}},
  booktitle = {ISNN},
  year = {2018},
  doi = {10.1007/978-3-319-92537-0\_45},
}

Loading related content...

Figure