Exact Bayesian bin classification: a fast alternative to Bayesian classification and its application to neural response analysis |
| |
Authors: | D Endres P Földiák |
| |
Institution: | (1) School of Psychology, University of St. Andrews, St Andrews, KY16 9JP, UK |
| |
Abstract: | We investigate the general problem of signal classification and, in particular, that of assigning stimulus labels to neural
spike trains recorded from single cortical neurons. Finding efficient ways of classifying neural responses is especially important
in experiments involving rapid presentation of stimuli. We introduce a fast, exact alternative to Bayesian classification.
Instead of estimating the class-conditional densities p(x|y) (where x is a scalar function of the features], y the class label) and converting them to P(y|x) via Bayes’ theorem, this probability is evaluated directly and without the need for approximations. This is achieved by
integrating over all possible binnings of x with an upper limit on the number of bins. Computational time is quadratic in both the number of observed data points and
the number of bins. The algorithm also allows for the computation of feedback signals, which can be used as input to subsequent
stages of inference, e.g. neural network training. Responses of single neurons from high-level visual cortex (area STSa) to
rapid sequences of complex visual stimuli are analysed. Information latency and response duration increase nonlinearly with
presentation duration, suggesting that neural processing speeds adapt to presentation speeds.
Action Editor: Alexander Borst |
| |
Keywords: | Classification Exact Bayesian inference Neural decoding |
本文献已被 PubMed SpringerLink 等数据库收录! |
|