A downward biased estimator of spike timing information

  • Authors:
  • Marcelo A. Montemurro;Riccardo Senatore;Stefano Panzeri

  • Affiliations:
  • Faculty of Life Sciences, The University of Manchester, The Mill, P.O. Box 88, Manchester M60 1QD, UK;Faculty of Life Sciences, The University of Manchester, The Mill, P.O. Box 88, Manchester M60 1QD, UK;Faculty of Life Sciences, The University of Manchester, The Mill, P.O. Box 88, Manchester M60 1QD, UK

  • Venue:
  • Neurocomputing
  • Year:
  • 2007

Quantified Score

Hi-index 0.01

Visualization

Abstract

We develop a new simple estimator of the spike timing mutual information between a set of static or dynamic stimuli and the elicited spike trains. Unlike the standard direct procedure (which provides upward-biased information estimation), this new method provides a downward biased (DB) estimator. Therefore, by using this new estimator in conjunction with the direct one it is possible to bound from both above and below the true asymptotic value of the mutual information. The downward bias property of the new method is useful in neurophysiological studies of neural codes because a finding of significant extra information in spike timing obtained with this new method will ensure that this additional spike timing information is genuine and not an artefact due to sampling problems.