Wesley, the reason that links coming out of the bias are not splittable originates with what I understood to be a canonical conceptualization of "bias," which...
Hey Wesley, sorry for the long wait on this question. I think the answer is that if you draw out the diagrams of the two alternative ways of splitting a...
Did the message, "Some thoughts about CPPNs," by petar_chervenski get deleted? I was going to respond to it but it seems to be gone. Petar, if you meant to...
Ken, I deleted it, because I couldn't edit it further and expand it with some other thoughts I had in mind. It was mentioned that I'll add to it, but it turned...
Hi everyone, I have been trying to evolve networks using rtNEAT which will learn a given target given the inputs, the target being a real number and the inputs...
2998
neat@yahoogroups.com
Mar 14, 2007 8:17 pm
Hello, This email message is a notification to let you know that a file has been uploaded to the Files area of the neat group. File : /Class...
Sure, glad to help. I've attached a zip of the modifications I've had to make in order to get it to compile on a modern Linux system (Fedora Core 5). The...
Hi Petar, within the next couple weeks we will finally be releasing new work involving CPPNs that will answer some of these questions, in particular the big...
Hi everyone, I am going to be speaking about NEAT software at an interesting workshop at GECCO 2007 called the "2007 GECCO Workshop on Open-Source Software for...
Hi I made my own C++ NEAT implementation, I called it TaoNeat. It main cool features are : * configuration with plain text file * OOP & plugin modular...
... Hello Alexandre, Is TaoNeat publicly available? I too have (mostly) written my own NEAT in C++, but am interested in other implementations. Thanks, Steve...
Not yet available ! I began to code it 2 months ago... There is no documentation excepted the comments, and there is also some part that don't satisfy me, from...
Ho, one more thing about my implementation : here it is an example of a configuration file. As we (maybe) can see, a configuration is just a serialized object...
Hi Stephen and Alexandre, if you are planning to release your systems to the public then I will be happy to mention them if you'd like, but please let me know....
... The reason mine's still considered incomplete, IMO, is because it comes nowhere near the performance you report in your original paper. I don't think I can...
Steve, not a problem, of course the pace you prefer to work is up to you. By the way, at what audience or venue would you be targeting your article? ken ... ...
... Audience: anyone who has read your papers, and wants to write their own NEAT implementation. My idea is to gather all of the information needed to do a...
Hi my name is Shane Ryan. I am researching using NEAT for time series prediction. I have been lurking in the group for the past few months. I've got a few...
Hi, I am in the midst of building a times series predicting framework based on Matt Buckland's windows c++ code. I've removed the visualization aspects. Made...
Hi, This isn't really specific to NEAT but I am not working with any other ANN technology right now. I was talking to some colleagues that are working in...
I have done the same thing under NEAT and rtNEAT using the C++ version. I have not been able to get particularly good results though from it. My scenario is...
Please, by all means. A Sourceforge project is a great idea. It would be nice to see a general purpose cross-platform C++ implementation in a centralized...
Hey Tom, Sounds like you are trying to use the ANN for decision making and not strictly time-series. But that doesn't really matter. IMHO it sounds like you...
OK I will try and set it up and have it posted on Sourceforge by the end of next week. I need to clean up the code and document it a bit. It won't be up to...
I am sure there are more than just a few of us who have had plenty of problems trying to get NEAT compiled and running. So posting something that is...
I second that. I also have experience wrapping Stanley's C++ code with Swig so I could help with porting the library to interpreted languages like Python. ...
Variable mutation rates are a valid idea. It has been tried a little to my knowledge, but not extensively. One useful thing might be to have variable...
I suggest you to read about MOGA (Mutation Only Genetic Algorithms) and AGA/LHDC/SHDC (Adaptive Genetic Algorithms). But it's not intended to increase...
Hi all Nice talk :D ! Instead of dropping the species that don't improve since a long time, giving them a luck with higher mutations rates since reasonable....
I've implemented a PacMan clone in which I've employed NEAT to evolve PacMan's AI. So far the results have been resonably good, and I hope to put the code in...
