, and how the will maybe one day we get there.
I decided that I take to write the blog of things not mathematical, since the amount of mathematics in my life as a scientist is, unfortunately, decreasing.
Today we had a very interesting article a group of scientists at Brandeis University.
Then, they took the baby ferrets and have them implanted with electrodes to measure neural activity. Then subjected them to several months to different types of visual stimuli: natural images, moving bars and white noise. (White noise can see below in all its glory).
They found that as the ferret was growing up, the statistics of spontaneous activity became increasingly similar to those of neural stimulation with the natural images, so that, after five months, the distribution of spontaneous activity was identical to the distribution stimulated with natural images.
This seems trivial, except that the images "natural" had only natural, so to speak, statistics. Have been using the trailer of the first Matrix film, such as samples of natural images, I guess just to make sure that only the general statistics of the image match those of the environment of the poor ferret.
This means, simply, that the statistics of spontaneous activity of the brain develops during the first phase of life (at least in the ferret) to a distribution in which an image has no pending representation from the point of view of global statistics of the neural network!
0 comments:
Post a Comment