Evolution of random synaptic weights of the hopfield associative memory : how chaotic trajectories turn into fixed point attractors?
Akira Imada, Keijiro Araki · NAIST Digital Library (Nara Institute of Science and Technology) · 1997
We apply evolutionary computations to Hopfield's neural network model of associative memory. We reported elsewhere that a fully connected neural network with random synaptic weights evolves to create fixed point attractors exactly at the locations of patterns to be memorized by a genetic algorithm. In this paper, we present the process of the evolution from chaotic behaviors to an associative memory.