How Does the Brain Organize Information?

Helmut Kroger
DOI: https://doi.org/10.48550/arXiv.0902.3418
2009-02-19
Neurons and Cognition
Abstract:Cognitive processes in the brain, like learning, formation of memory, recovery of memorized images, classification of objects have two features: First, there is no supervisor in the brain who controls these processes. Second there is a hugh number of neurons (10^{6} to 10^{10}) involved in those cognitive tasks. For this reason, the search of understanding cognitive processes uses models built from a large number of neurons, but very much simplified neurons. The so-called neural networks have been quite successful in describing certain aspects of brain functions, like the mechanism of associative memory or recently the prediction of epileptic seizures. At hand of the Kohonen network we discuss the treatment of information in the brain, in particular how the brain organizes such information without supervisor. Recently, networks of small-world and scale-free architecture came into focus. There is evidence indicating that the brain (cat cortex, macaque cortex, human brain) uses such connectivity architecture. Tasks like treatment of information, learning and classification take advantage of such scale-free and small-world connectivity and thus play a potentially important role in self-organization of the brain.
What problem does this paper attempt to address?