Some Computational Principles derived from Cortical Neuroanatomy
演題番号 : S1-1-1-2
Rodney James Doulgas:1
1:Institute of Neuroinformatics, Univ of Zuerich, Zuerich, Switzerland
Over the past decades neuroanatomical investigation of the neocortex has yielded a wealth of information concerning the various types of neurons and their connectivity. However, progress on understanding how these circuits operate to support perception and cognition has been relatively slow. Indeed, at present it seems that very large scale simulations of neuronal networks stand proxy for real algorithmic understanding of cortical processing. In this talk I will outline some computational principles and constraints that can be extracted from quantitative cortical neuronanatomy, and show how they can be implemented in neuromorphic electronic hardware.