Title: Inference for first passage times of the Feller process
Authors: Satish Iyengar - University of Pittsburgh (United States) [presenting]
Abstract: The Feller diffusion process has linear drift and a state dependent diffusion coefficient that vanishes at zero. Earlier studies have shown that it provides a better fit for neural activity than the Ornstein-Uhlenbeck under certain conditions. We describe inference based on maximum likelihood for this model when the available data are spike trains rather than the neurons subthreshold voltage traces.