E-M ALGORITHMS FOR ESTIMATING PARAMETERS FROM SINGLE-MEMORY MARKOV POINT-PROCESSES HAVING A MULTIPLICATIVE INTENSITY.

The probability that a neuron in the peripheral auditory nervous system discharges during some specified time interval depends on both a stimulus function and on effects due to earlier discharge events. With the objective of estimating parameters related to the stimulus function which are free from the distortion introduced by effects due to previous discharges, a maximum-likelihood estimation scheme has been developed. We have chosen a multiplicative intensity model to describe auditory-nerve discharges in which the intensity of the process is stochastic and the product of two functions: one an unknown deterministic function corresponding to the stimulus, and the second a stochastic function which is on the history of the process. Only abstract is given.