SOME OPTIMAL STOCHASTIC CONTROL PROBLEMS IN NEUROSCIENCE — A REVIEW

Nervous systems are probability machines and, as such, modeling their activities should incorporate stochastic processes. In this review, we present two examples of optimal stochastic control problems with an analytic methodology on how to find optimal signals. The first example deals with neuronal activity and the second example is concerned with a higher level task: arm movement. In both cases we find optimal signals for particular tasks and find our results in agreement with the experimental Fitts Law.