Maximum likelihood estimation of a stochastic integrate-and-fire neural encoding model

L Paninski, J Pillow, and E P Simoncelli

Published in Neural Computation, vol.16(12), pp. 2533--2561, Dec 2004.
© MIT Press
DOI: 10.1162/0899766042321797

Download:
  • Reprint (pdf)
    • We examine a cascade encoding model for neural response in which a linear filtering stage is followed by a noisy, leaky, integrate-and-fire spike generation mechanism. This model provides a biophysically more realistic alternative to models based on Poisson (memoryless) spike generation, and can effectively reproduce a variety of spiking behaviors seen in vivo. We describe the maximum likelihood estimator for the model parameters, given only extracellular spike train responses (not intracellular voltage data). Specifically, we prove that the log likelihood function is concave and thus has an essentially unique global maximum that can be found using gradient ascent techniques. We develop an efficient algorithm for computing the maximum likelihood solution, demonstrate the effectiveness of the resulting estimator with numerical simulations, and discuss a method of testing the model's validity using time-rescaling and density evolution techniques.
    Related:
  • Superseded papers: Pillow03bMaximum Likelihood Estimation of a Stochastic Integrate-and-Fire Neural Model
    by J W Pillow, L Paninski, and E P Simoncelli
  • Previous conference publication of this work: nips-03Maximum Likelihood Estimation of a Stochastic Integrate-and-Fire Neural Model
    by J W Pillow, L Paninski, and E P Simoncelli
  • Application to retinal data: JN-05Prediction and decoding of retinal responses with a probabilistic spiking model
    by J W Pillow, L Paninski, V J Uzell, E P Simoncelli, and E J Chichilnisky
  • A more general book chapter on characterizing neural responses: Gazzaniga-03Characterization of neural responses with stochastic stimuli
    by E P Simoncelli, J Pillow, L Paninski, and O Schwartz
  • Description of biasses in reverse-correlation (spike-triggered average) estimates of the LNP model: cns-02Biases in white noise analysis due to non-Poisson spike generation
    by J W Pillow and E P Simoncelli
  • Online Publications