The brain is often considered a noisy system. Trappenberg argues that variability is deliberate and necessary to guide human behavior in uncertain environments. While this has been captured to some extend in Bayesian frameworks, he also argues that we need to go beyond this abstract framework and consider practical implementations and behavioral implications. In this context he would like to outline some of his research that spans cellular, network and system level examples of computational neuroscience. In particular he would like to present a neural field model of an arm controller that represents a neural population implementation of a Bayes (Kalman) filter. He would also like to show that synapses are probabilistic and how learning an appropriate probabilistic response could give rise to advanced systems. 

 

© UC Irvine School of Social Sciences - 3151 Social Sciences Plaza, Irvine, CA 92697-5100 - 949.824.2766