Hi All, Copied from my LJ post at http://the-locster.livejournal.com/18585.html ... [...picbreeder explanation cut...] Given a network it is possible to tweak...
Ken, Jason and David, Thanks for the new paper. This is very powerful stuff. Ken ============================= Kenneth A. Lloyd CEO and Director of Systems...
I am thinking of applying for a grant to apply NEAT to video games. I am interested if anyone has ideas on applying NEAT to non-first person shooter (FPS)...
Hi Shane, Well there is already one application of NEAT to a game, Neuro- Evolving Robotic Operatives (NERO). For this game a real-time version of NEAT was...
It isn't very sophisticated, but I have a PacMan clone in which the PacMan AI is controlled via NEAT. http://sourceforge.net/projects/aipac/ -Kirk On Sun, Sep...
Hey all, HyperNEAT 2.4 is out. This version includes bug fixes to the Checkers and FindCluster experiments and performance increases. I really appreciate all...
Hey Kirk, Cool demo. But it keeps crashing on me. I've gotten it to run the actual Pacman a couple of times around the maze. But generally when I start it up....
Hi Kirk, If it helps I notice the problem disappears when I reduce of ghosts to zero. Also the "Gens no improvement" statistic often shows values that exceed...
Hmmm....interesting on the freeze up. I haven't seen that one before. Make sure to read the README file that I posted on SourceForge and that should be ...
You probably won't be interested in this post since you're not interested in using expert knowledge. But what if you used Bayesian probabilities as a form of...
Peter, we do need to add to Picbreeder a display of genome complexity for each image. That would certainly answer some questions. I am going to see if we can...
We definitely need to work on that tool, it would be very educational. I hope we can find the time. ken ... and meet ... the tool ... is any ... a good ... ...
Hello "Along these lines, I have rarely seen a cellular growth model produce anything more than a blob or an amorphous shape without clear parts." If you...
Alexandre, thank you for bringing up reaction-diffusion systems. You make a good point- my statement was too broad, in part because "cellular growth model" is...
<nitpicking mode on> I don't think that the capacity to build sharp frontiers is a real issue. A good old sigmoid applied to a linear combination of the...
I'd like to point out that CPPNs can be combined effectively with cellular growth models. Because the cells need the answer to the question "Who am I?", the...
To follow the details of the argument, I think the issue is not so much with whether or not such systems can build a sharp frontier but with the structure of...
Guten Tag To make a reaction-diffusion system, you have to pick 1) chemicals diffusion coefficients 2) a Laplacian operator (how to compute the variation of...
Okay I understand the zero-border idea now, thanks for the explanation. But that does not seem to answer the question of how the reaction-diffusion system...
Peter, you make a good point that there is a challenge in interpreting arbitrary CPPN patterns as distinct physical objects. I think ideas like yours are...
Hi Ken, I was looking for an algorithm to develop animal morphologies (generally any kind of solid physics objects) with CPPNs. I found that writing a...
Hi all, I'm interested in applying NEAT to the development of a controller for a simulated biped robot. So to get my feet wet, I've programmed NEAT in Matlab....
Joe, I'll give it a shot. First, I suggest taking a look at the question, "How should I test my own version of NEAT to make sure it works?" in the NEAT FAQ...
Peter, I understand your ideas and they have good motivation. There is a tricky issue with interpreting the pattern output by a CPPN as the morphology of an...
I seem to have gotten all the dependencies sorted out, all that remains is what seems to be language syntax differences between compilers (or versions). Which...
Hey Oliver. I run ubuntu myself and it works fine, but I use a custom version of wxwidgets. I'm going to look into what might be causing your problem with the...
Hey all, This issue has been resolved (some of Oliver's CMAKE variables were set wrong) but has uncovered another issue: The MPI HyperNEAT extension has a...
Hello all- A while back we had a discussion about whether natural selection is good at choosing its own parameters. Many people assume that it would be...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
