Hi Paul, Yes, you are right about that pruning of neurons is probably not a biological performance issue. It was just some hypothesis I thought of. You are...
I am still working with Buckland's NEAT for Windows code. I have successfully muted a genome following his convention. Now I want to do the speciation and...
Hello NEAT group, I've been working with jNeat and I've come to a bit of confusing output, so Nate Kohl pointed me this way. Anyways, what happened was I was...
Hi Aidan, The original reference implementation of the 'hard' learning problem is available (with source) from: http://eplex.cs.ucf.edu/software.html (In the...
John, that sounds like a potential bug. Can you find in the code where the mutate_add_link and mutate_add_node methods are called? You should be able to see...
I probably should have done that from the start. Anyways after searching around for a while for different variables to output and where, while trying to figure...
I did what I should of done prior to my last posting. I went back and reread the NEAT chapter from Mat Buckland's book "AI Techniques for Game Programming." ...
I was wondering if there was every any discussion of releasing GAR on the Xbox 360? I recently stumbled upon some information regarding submitting games to the...
Yes, in fact we are hoping to port GAR to Xbox. Because GAR is written in XNA, it should be feasible. There are some tricky issues between PC and Xbox, but...
Quick question. What's the prevailing opinion on whether CPPNs should be recursive or feedforward-only? Also there were discussions some years back about...
Hi Colin ... I've recently implemented an algorithm that find this longuest path. It is written using boost::graph for my own framework (sferes2) but you could...
Colin, it is probably not a critical decision whether CPPNs are recurrent or not (at least for the purposes of HyperNEAT), but it does have implications. In...
Colin, there is indeed a fast way to compute feedforward activation, but it does not actually involve computing the longest path. Rather, it entails recursing...
2009/11/3 Ken <kstanley@...> ... [...] ... Yes I see. It's like evaluating an expression tree but with an extra rule for handling the fact it's not a...
2009/11/3 Ken <kstanley@...> ... [...] Thanks. From reading your response my instinct is to use feedforward as the default/baseline for CPPNs and to...
Hi Colin, ... Yes, this is exactly the second approach listed in the DAG part (i.e. "Using shortest-path algorithms") since a feed-forward NN is obviously a...
Hey Colin, CPPNs, being an indirect encoding, allow for small changes in the genotype to have profound effects on the resulting substrate. While this is what...
Can anyone please recommend some HyperNEAT documentation, a tutorial, diagram, some clue, or anything that does not mean reverse engineering the comment-less...
Andrei, which version of HyperNEAT are you interested in and what references have you looked at so far? We can potentially improve the documentation based on...
I should have specified that I was talking exclusively about the HyperNEAT 2.6 C++ implementation. I read some of the publications and the ideas are...
Andrei, if you want to learn the underlying NEAT method by looking at c++ code, the best I have seen is Dr. Kenneth Stanley's rtNEAT c++ code. It is fairly...
Dear all, Following the IROS workshop on Evolutionary Design of Robots held in october 2009, during IROS, we are starting a new mailing list regarding general...
Nicholas, Thanks for the notice. Likewise, be aware of HART (Human, Agent, Robot, Teamwork). There is substantial research and application of evolutionary,...
Buckland's NEAT for windows code uses the Adjusted Fitness to calculate Span Amount. Is this incorrect? In CSpecies::AdjustFitnesses(), for each member of a...
Buckland's NEAT for windows code uses the Adjusted Fitness to calculate Span Amount. Is this incorrect? In CSpecies::AdjustFitnesses(), for each member of a...
Yes I would advise starting out with NEAT to get familiar with it. However, we can also look into improving the understandability of the HyperNEAT C++...
Hey Andrei, It's true that the HyperNEAT experiments are poorly documented and can be confusing. This is largely because the HyperNEAT algorithm itself has...