Paradoxical Relationship between Output and Input Regularity for the FitzHugh-Nagumo Model

Abstract. We examine the effects of changing the coefficient of variation (CV) of the inter-stimulus interval on the CV of the output interspike interval (ISI), using constant magnitude, supra-threshold pointprocess stimuli of the membrane potential variable in the FitzHughNagumo model. The coefficient of variation of the input is changed within the context of a displaced exponential distribution. We find that for some values of mean inter-stimulus interval, CV of ISI has an inverse relationship with input coefficient of variation, whereas for other mean stimulation rates, CV of ISI increases with input coefficient of variation. Over a wide range of input regularity CV of ISI is approximately constant, and is less than 0.4.

• Type: Conference
• Year: 1999
• Where: IWANN
• Authors: Stuart Feerick, Jianfeng Feng, David Brown
• Keywords: Input Coefficient, Inter-stimulus Interval, IWANN 1999, Neural Networks, Output Interspike Interval