STOCHASTIC PROPERTIES OF NEURONS WITH SHUNT TYPE INHIBITION.

Stochastic properties of neurons with shunt type inhibition which have Poisson pulse inputs are considered. The shunt type inhibition lowers the neuron potential in the form of division. In this case the neuron model becomes a system with state dependent noise. The transition probability density function and the first passage time density function of the neuron potential are important. These quantities satisfy Kolmogorov's equation and the law of probability conservation. However, analytical solution cannot be obtained except in very simple cases and hence the numerical method is used together with the analytical one. The results calculated for shunt type inhibition are compared with those for subtractive inhibition.