作者: Emmanuel Daucé , Frédéric Henry
DOI:
关键词:
摘要: This paper presents learning simulation results on a balanced recurrent neural network of spiking neurons with simple implementation the STDP plasticity rule, whose potentiation and depression e ects compensate. The synaptic weights delays are randomly set activity, which is combination an input signal feedback, initially strong irregular. Under static stimulation, process shapes initial activity toward more regular synchronous response. response speci c to this particular stimulus: has learned select by synchrony one arbitrary stimulus from random stimuli.