Title :
Mean-field theory of the three-level associative memory and its synthesis
Author_Institution :
Inst. for Network & Syst. Theory, Stuttgart Univ., Germany
Abstract :
The single layer neural network with feedback and discrete-time, synchronous dynamics serves as an associative memory for bi-valued patterns. A three-level function instead of the normally used signum-function as output function of the neurons decreases the regions of attraction of the spurious states considerably. The mean-field theory is applied to this kind of system. Results of the theoretical investigation explain its behavior and provide expressions for the choice of the function parameters. The theory is in agreement with results obtained from simulations
Keywords :
content-addressable storage; recurrent neural nets; associative memory; bi-valued patterns; discrete-time synchronous dynamics; feedback; mean-field theory; output function; signum-function; single-layer neural network; spurious states; three-level associative memory; Associative memory; Equations; Hardware; Information retrieval; Network synthesis; Neural networks; Neurofeedback; Neurons; Recurrent neural networks; Symmetric matrices;
Conference_Titel :
Neural Networks, 1995. Proceedings., IEEE International Conference on
Conference_Location :
Perth, WA
Print_ISBN :
0-7803-2768-3
DOI :
10.1109/ICNN.1995.488888
