Output details
15 - General Engineering
University of Warwick
Maximally informative stimuli and tuning curves for sigmoidal rate-coding neurons and populations
This paper was the first to find closed-form expressions for the information-optimal neural tuning curves that maximise Shannon mutual information and it was the first to avoid the (standard) assumption of Poisson statistics. The work was funded by the EPSRC grant “Multiplicative and fractal noise coding for cochlear implants” as part of our ongoing efforts to understand the coding of auditory stimuli by the human ear. This publication resulted in an invitation (fully supported financially) to present the paper at the International Workshop on Statistical-Mechanical Informatics (Kyoto, Japan, Sept. 2009). Contribution to the field is articulated in work citing this paper.