The evolution of overlap which is easily obtained in Hopfield neural network with continuous hyperbolic tangent thresholding function is equivalent to that in corresponding two-state neural network with finite tempera...The evolution of overlap which is easily obtained in Hopfield neural network with continuous hyperbolic tangent thresholding function is equivalent to that in corresponding two-state neural network with finite temperature.The attractors of two simple cases,the Hopfield model with independent random patterns and pair correlated patterns,are analyzed.展开更多
文摘The evolution of overlap which is easily obtained in Hopfield neural network with continuous hyperbolic tangent thresholding function is equivalent to that in corresponding two-state neural network with finite temperature.The attractors of two simple cases,the Hopfield model with independent random patterns and pair correlated patterns,are analyzed.