|
About
NeuroCOLT
Papers
Archive
Books
info@neurocolt.org
|
NeuroCOLT
Technical Report NC-TR-00-064
Paradigms
for Computing with Spiking Neurons
Wolfgang Maass
Abstract
Spiking neurons differ in essential aspects from the familiar computational
units of common neural network models, such as McCulloch-Pitts neurons
or sigmoidal gates. Therefore the question arises how one can {\em
compute} with spiking neurons, or with related computational units
in electronic hardware, whose input and output consists of trains
of pulses. Furthermore the question arises how the computational power
of networks of such units relates to that of common reference models,
such as threshold circuits or multi-layer perceptrons. Both of these
questions will be addressed in this paper.
Download
Compressed Postscript
|