The accuracy of the population vector estimate in networks of integrate-and-fire type neurons using stationary and

摘要

We study the accuracy of a population vector method. We consider a network of spiking neurons where each neurion responds preferentially to a particu- lar direction of the stimulus. We show that the accuracy of the population vector depends on the coupling weights between neurons in the population. Some types of interactions produce more accurate estimations for stationary stimuli whereas other types are better for transient stimuli