EFFECT OF CORRELATED FLUCTUATIONS OF SYNAPSES IN THE PERFORMANCE OF NEURAL NETWORKS

We consider stochastic neural networks in which synaptic intensities r apidly fluctuate-around means given by a learning rule-competing with neuron activity. Each snapshot of synaptic intensities contains the ne uron-neuron correlations in one of the stored patterns chosen at rando m. The result is apparently noisy behavior which induces robustness, i ncluding improved associative and pattern recognition processes. The m ain result here might apply to biological systems that exhibit fluctua ting patterns of synapses.