I am looking for a open-source framework to implement
multi-agent simulations. I am currently using my own
program - it is written in Lisp, which means it's
limited and hard to distribute in/share with the community.
What I am looking for, essentially, is a way to put
standard recurrent networks into feedbackloops with
each other, where the connnections between the
different seperate networks (i.e. agents) are adaptive
in the same way as connectionns within the standard
networks. In other words, embed recurrent networks
within a larger, extended recurrent network.
Short description of my current model can be found here
or here http://homepage.ntlworld.com/dirk.vanrooy/
It looks like Joone would be perfectly suited for this
type of thing. Anybody else working along those lines?
Log in to post a comment.