... The size of the viewspace can evolve to some extent, but not very much. ... The viewspace decides where it wants to look. It basically "looks around" the...
Hi all... We've had quite a few people join up in the past week...if you haven't already, please jot us a note introducing yourselves and letting us know your ...
Welcome, My name is Ugo Vierucci and I am interested of AI to some years. After to have read the characteristics of NEAT they are decided to me to create one...
45
Chris Christenson
cc57@...
Sep 2, 2003 4:15 pm
Hello, My name is Chris Christenson I am a Masters student at Texas State. My thesis has to do with evolving neural networks and so I plan to use NEAT as my...
Hello, My name is Jon Crosby and I decided to join out of curiosity after reading the some of the available NEAT literature on the net. I hope to be able to ...
Hi all, Chris mentioned that he's working with NEAT for his Master's thesis...which implementation do you plan to use? Or do you plan to write your own? There...
50
Chris Christenson
cc57@...
Sep 3, 2003 6:13 pm
Hello, I plan to use the Windows C++ version. C++ because that is the language I am best at, and since I plan to alter the code a little that seemed the best...
One thing people may not realize is that although I originally wrote my version of NEAT in C++ for Linux, it can be compiled in Windows (sacrificing the...
Hey Everyone, My names Jonathon Vincent, I have reciently completed Computer Science at RMIT University in Melbourne. In my final Semester I worked on a...
... <snip> ... For our implementation, Philip and I are working on a browser-based interface, basically an HTML form in which you select the evolutionary...
54
Chris
cc57@...
Sep 4, 2003 2:55 pm
QT provides some nice GUI tools, and it is cross platform compatible. That might be a good way to go. -Chris C ... From: Kenneth Stanley...
Some of you may be interested to know that Christian Mayr produced a version of NEAT for Matlab, which is now available through my homepage and the NEAT Users...
... This sounds like it will be really cool. Will graphics also be displayed in the browser during runtime? What if you ran an experiment for example...
... The approach we're taking is to have any additional graphical information pop up in a new browser window. For something like an XOR run, you would have ...
Since the group is becoming fairly large, this is probably a good opportunity to see what the general opinion is about an important question: What do you...
One could use NEAT for motion evolution as the UK company did: www.naturalmotion.com very interesting stuff that NEAT neural network can be used! Jianjun...
Hi All, I am one of the guys meeting every two weeks with Derek and Philip. At present, I have read the papers Derek has sent me on NEAT. And I have read the...
Hello All ! I'm Umer, from Pakistan. I completed my BS in Computer Systems Engineering from GIK Institute, Pakistan this May. During my last two semesters, I...
... Welcome to the group, Umer. Your experiment sounds quite interesting. ... So what was the body design of your robots? Were they entirely simulated, or...
Hey Umer, Concerning your predator and prey stuff, I agree the the reason for the failure to evolve effectivly was due to the robots not being able to tell the...
Hey all, Many times during our attempt to implement NEAT, we've e-mailed Ken with various questions for clarifications. I was hoping this group could help ...
Joseph and I talked a bit on this group about implementing NEAT with indirect encoding, taking a modular approach. And I know Ken has lots of thoughts on...
... I should clarify a couple things about the way I handled speciation. There are two important properties of a species: 1) How many members it contains in ...
Hi Umer, ... Sounds like a very interesting experiment. Thanks for sharing it. A few questions: Do you have any pictures or animations from your work? Also,...
... I'm not sure if I've mentioned this yet, but one of the things I'd like to see is a system that evolves the nn developmental system rules along with the...
... Why would they need to be logically decoupled? The DNA string could could contain instructions for building particular structures, and developmental rule ...
... Yeah, sure, and this is the way its normally done I think. I want to split them up in order to take advantage of coevolution. Also this way different ...
... split ... different ... chromosomes. ... So is your idea mainly to coevolve DNA "starter strings" in one population and "rule sets" (or something...
