System identification based on point processes and correlation densities. II. The refractory neuron model (Q1111957): Difference between revisions

The results presented by the author, \textit{G. Hesselmans} and \textit{P. Johannesma} in part I of this paper, ibid. 80, 143-171 (1986; Zbl 0613.92013), are further developed. The studied neuronal model is composed of three parts: two of them the same as in the previous paper and additionally a linear filter that forms a feedback of the output point process. The feedback causes the output point process to be a self- exciting one. An example of the feedback aiming to describe a dead time after neuronal firing and exponential recovery is considered. The auto- and cross-correlation densities for the model neuron are mainly derived in the case given by the special example. The simulations of the model illustrate the results achieved.