| We studied the history dependent processes for the spike train when the spike train is in the weak stationary boundaries based on a novel test strategy for finding the stationary periods of neuronal activity. Furthermore, to incorporate serial interval correlations in a framework of stochastic point processes, we developed an ad hoc technique for generating and modeling serially correlated event processes that can be efficiently simulated. we provide the maximum likelihood estimators for the parameters of the model under the assumption of log-normal inter spike distribution. However, the model could be modified asymptotically to any given distribution which is well fitted to experimental data. |
|