This post is not strictly about NEAT and is general to the field of neuro-evolutionary systems but is relevant nonetheless, and since I don't know many good...
... Peter, I was curious that you said you are now not sure if your method is the right way to evolve substrates. Why are you having second thoughts? Do you...
Ken, Peter, Have you tried separating patterns of growth from patterns of evolution.? Specifically, are ontogenic patterns a separate concern from phylogenic ...
Antoan, I think your comments and questions are entirely appropriate for the group. Issues of general interest to neuroevolution are often discussed here and...
... Yes, I would. :) This topic is highly relevant to this forum. ... Well, there are different levels of modularity as well. A neuron is a reusable module. At...
Hi all- A quick question for Ken, Jason and Dave. I notice that most of the parameter settings (e.g. MutateLinkWeightsProbability) were the same when you guys...
Hi Ken, Well I did have second thoughts, because I was uncertain. I need to explore the possibilities of the method more before I have any hard conclusions....
Hey Jeff, I understand the concern with the mutation probability. I believe the reason that NEAT cannot solve the boxes problem in training is a credit ...
Jeff, a few other thoughts on the issue... Note that I believe the only time P-NEAT comes up is in the boxes domain. None of David's experiments involve...
Hello everyone, I have finally found the mistake I made. Between samples, the arm's speed in the simulator was not set back to 0, resulting in the arm starting...
Matthias, glad to hear it's been solved. About your new question, I would suggest avoiding randomness when possible, even if it means a bit more evaluation....
... wrote: Peter, I think a key issue is the one you raise about the raw CPPN not ... You make a good argument, and I understand the concerns, but I think this...
Ken, I didn't completely understand what you mean by "phylogenetic patterns?" Does that refer to a phylogenetic tree or does it refer to perhaps how different...
Hi Ken- Out of curiosity, how is the picbreeder site coming? It appears to me that there is a sustained level of activity, but I was wondering if you guys have...
Hello- I think this email that I received from a NN email list is related to this subject, and that you mind find its content interesting. I would also be...
Hello, Is there a publicly available Perl version of NEAT (or any of the variants)? If there's some existing code out there, I would probably just use that...
Ben, I do not know of any Perl-based NEAT implementations, so yours would likely be the first. So it would probably be appreciated by those who prefer Perl. ...
Picbreeder usage is fairly consistent over the long term with occasional spikes caused by modest media exposure. We have a few regulars who tend to produce a...
Ken, Moose is a new object framework for Perl that simplifies and standardizes a lot of things. The archives show John Arrowwood <jarrowwx@...> said "I'm...
Hi Ken, Jeff, Regarding the merchandise it probably makes sense to get the opinion of a marketing expert - contrary to popular belief they're not all useless...
Colin, thank you for the idea. You may indeed be right that a page of selected merchandise would work better, that's a good thought. The whole merchandise...
Cesar, How is your 1.0 implementation progressing? If there is anything you would like checked/worked on then let me know; would be great to have it done for...
Thanks for this very detailed response Ken. I am glad to hear people continue to use the site. I am planning on buying some of the custom products at some...
Jason and Ken- Thanks for your responses. I think comparing HyperNEAT and P-NEAT is very instructive, because it allows us to test the benefits of generative...
Jeff, you make a legitimate critique: The study could be improved with an extensive parameter sweep for P-NEAT. As a routine matter, we do generally try some...
Hi Ken- I agree with you. Generalization and scaling are very cool and important results from that paper, and they have nothing to do with having the proper ...
Hi all, If you consider a genome instance to be a point in a multidimensional space then perhaps Euclidean distance is a sensible/reasonable metric to use when...
Just had a few more thoughts on this one. Consider a family tree of genomes where each mutation is a essentially a choice to go down one branch of the tree....
Colin, First, thanks for your contributions. You've been a great help in developing my understandings of NEAT. (I hope that is a compliment). Are you saying...
Hi Ken, ... Thanks. ego boosts always welcome :) ... I'm saying that the current compatibility scheme in NEAT is effectively using the L1 norm (Manhattan...
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.
