Message 180 - Posted 15 Oct 2011 22:06:57 UTC
Dear friends,
After analysing the results from this Summer, the model has been found not very conclusive but still intriguing.
The best result was found in the test number #464108. This solution allowed the model to evolve for some time with no metal activity extinction in a very irregular but interesting pattern.
Maybe the general model using inhibition and excitation is not able to reach a periodic pattern but only a chaotic one. This is unexpected and, somehow, a drawback with respect to my expectations, but could allow us to tune the model to follow a real EEG.
Here I have a number of new WUs to try. They are, somehow, special. They are designed to explore the upper limit of the network connectivity, and this implies that they are going to require more RAM, between 4 and 16 GB, so get prepared. As I wrote in a past post, is like having a look to the dark side of the Moon. Never thought I would be able to model them, but with your support it will be possible to see what is in there.
In the following days I will make available small sets of WUs to ensure that they run smoothly. Also I will try with different ratios of redundancy to study how the credits work with the new computation times.
I count on you, as always, and thank you for your support.
I'll be right here for you.
Regards,
Javier.