A Model for Real-Time Computation in Generic Neural Microcircuits

Wolfgang Maass, Thomas Natschläger, Henry Markram
2002 Neural Information Processing Systems  
A key challenge for neural modeling is to explain how a continuous stream of multi-modal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real-time. We propose a new computational model that is based on principles of high dimensional dynamical systems in combination with statistical learning theory. It can be implemented on generic evolved or found recurrent circuitry. © PI ), the firing rate of spike trains 3&4 ( ©
more » ... ), the sum of © QI and © # in an earlier time interval [ § -60 ms, § -30 ms] ( © F ) and during the interval [ § -150 ms, § ] ( © 7 ), spike coincidences between inputs 1&3 ( © SR ¡ ¨ § ¤¥ is defined as the number of spikes which are accompanied by a spike in the other spike train within 5 ms during the interval [ § -20 ms, § ]), a simple nonlinear combinations © UT and a randomly chosen complex nonlinear combination © WV of earlier described values. Since that all readouts were linear units, these nonlinear combinations are computed implicitly within the generic microcircuit model. Average correlation coefficients between targets and outputs for 200 test inputs of length 1 s for
dblp:conf/nips/MaassNM02 fatcat:xg4b5c6zwvbbdefzmwjvrre6sa