- Welcome to the NEAT Users Yahoo Group.
Presumably you're here either because you use NEAT or
are interested in using NEAT (or you created NEAT).
You already know the purpose of this group and list,
but I thought I'd introduce myself and Philip Tucker,
and give you an idea of how we got interested in NEAT,
what we're doing right now, and where we're going.
My name is Derek James, and my background is not in
either Artificial Intelligence or Computer Science.
It's in education. I graduated from U.T. Austin in
1993 with a B.A. in English and teacher certification
in Math, Science, and English. But I've always
considered AI an avocation, and I've read extensively
on the subject since I was in high school. Philip and
I met at Baylor University in 1989, where he was
working on an undergraduate in CS. He went on to get
a Master's, also in CS.
We've been friends since then, and had discussions
regarding AI, but never really took any steps to
actively develop AI. That changed last year. Philip
enrolled in an introductory neural network class here
at the University of Texas at Dallas. I audited the
class. We met two other guys, one who was nearly
finished with a Master's in Cognitive Science from
UTD, the other a developer who was working on AI in
his spare time. The four of us started a bi-weekly
discussion group centered around using neural networks
and genetic algorithms.
Like many AI researchers, our group discussed a
variety of domains, but we were primarily interested
in the domain of classic board games, specifically Go.
Around the end of 2002, I began to read published work
in the field. At one of the group member's urging, I
read Fogel's book, Blondie24, regarding evolving ANNs
with fixed topologies as part of a Checkers-playing
I read about the work done at UT Austin over the past
decade or so. I read first about SANE, then found the
paper on NEAT. It was exactly the sort of approach I
was looking for. I showed it to Philip, and we
decided that it was a technique we would be interested
in experimenting with.
We looked at existing implementations of NEAT, but for
a variety of reasons we decided to implement our own
version. It is built around two existing open-source
JOONE (Java Object-Oriented Neural Engine)
and JGAP (Java Genetic Algorithms Package)
There are a number of features in these packages,
especially JOONE, which we are not using in our
initial implementation, but which might be useful
We wanted to implement NEAT so that it would be easily
adapted to a distributed computing environment (we
believe that a high-input/output domain like Go will
require such an environment). We also wanted to
implement it so that it would run through a browser
interface. This version will also persist a number of
graphical diagnostics, including graphical
representations of the evolved neural nets,
evolutionary progress, etc., in the XML-based
graphical language, SVG.
As of this writing, the core NEAT engine is
implemented and working, though we have not
incorporated speciation yet. We have tested it so
far, without speciation, on XOR, and gotten it to
converge in multiple trials. It will probably be a
few more months before we have a stable,
fully-implemented version of NEAT, though.
At that time, we plan to initiate an open-source
project on SourceForge and make the source freely
available to anyone who is interested. We would also
like to have two or three demo tasks available for our
implementation before we release it.
As far as our own research path, we would like to
explore competitive coevolutionary techniques,
specifically applied to game domains. Our first game
domain will be Tic-Tac-Toe, and we hope to scale up to
GoMoku (5-in-a-row on a 13x13 board), and then Go. We
also hope to explore some form of indirect encoding to
exploit the inherent symmetry in such game domains.
So that's who we are, where we're at right now, and
where we hope to be in the near future. If you join
the group, please consider introducing yourself
(though you don't have to be as verbose as me) and
letting the group know how you're either using NEAT or
would like to.
Thanks for joining, and we'll see you around...
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
- Hi, I'm Kenneth Stanley. I think Derek had a great idea starting this
group. I have been in contact with a lot of people who are working
with NEAT independently, and it makes sense to have a place where we
can pool our experience to help each other out. The group can also
facilitate the discussion of ideas, both for applying NEAT and
extending it. I look forward to participating!
As for introducing myself, many of you already know who I am, so I
won't go too long on the details, but for any who don't, I am
completing my Ph.D. in computer science at the University of Texas at
Austin. I am hoping to become a professor after graduation so that I
can continue research into evolving increasingly complex neural
networks, as well as other structures. I don't know where I will be a
professor- it's not an easy job market! In the future, I hope to lead
research towards the next generation of neuroevolution systems
building on NEAT.
- Hi Group:
My name is "Germán" it sounds like "her man" (but I'm not a
playboy...unfortunately), I'm from Spain and I've been dedicated to AI as a
hobby since 15 years ago. I side of my job is as "Engineer Consultant"
(designing systems for the industry) the other side is "Business Consultant"
(improving organization performance).
I've a lot interest in NEAT because its ability to find simple solutions,
its reduction of searching space through complexification, its ability to
keep several possibilities open in the solution space through speciation.
Some of my thoughts about ANNs:
- Now I'm studying the Meta-learning or the process where the systems
learn to learn, I think the future are in that kind of systems, why?:
- I think an <<Intelligent>> Systems is such a system that not only is able
to solve a task, as well it is able to use all information it has learned
until a time to accelerate its learning speed rate in the immediate future.
In this way I've seen some success experiment with NEAT and 'Go' game
- But, what is that experience?. They are not only useful parts to be
reused; as well they are modifications in ANN structure searching space. To
say: if exist an ANN Space of Solutions (N) that is projected in the
Solution Space (inputs/outputs) (S) and there are n tasks (r1, r2.rn) that
my system has been able to solve, my system must not be only a mix of
Neurons in the N space and having the ability to solve such all tasks (<=n).
My systems must be able to deform dynamically the searching space (inside N)
using the previous learned information to be able to find quickly new
solutions for task r_n+1 (Obviously if such task has a relationship with
- So, what have to change dynamically in a GA to look for the solutions in
the right places?.... Mutation and Crossover. Imagine this yahoo group is a
GA where we are "solutions", everybody has his own knowledge about ANNs.
Each message we send is a "crossover" process where we are sharing
experience and knowledge, however we don't a fix or random crossover, we are
intelligent "solutions" and thus, we are able to incorporate to our
knowledge the new part, we do a "intelligent crossover" or "probabilistic
crossover" based in the new information and in our previous experience,
besides when our solution "mutates" because we try something new in a ANN,
we try to maximize our probability to get a success result. (Normally in an
- I think indirect encoding as DNA is a way to reduce the searching space,
but DNA doesn't work as puzzle book to build life. DNA has evolves too. It
has got through evolution mechanisms to manage other part of DNA and
probably to mutate in the right way depending of natural selection
competitiveness; in fact, recent studies have demonstrated that since the
beginning of evolution there were immutable DNA pieces sharing by all alive
beings. Probably under that pieces are the secret mechanism to evolve in a
So...my actual studies are about "Indirect encoding": not only encoding ANN
structures (I think to use it only is a lost war because GAs need something
more to deal with such complex searching spaces), as well encoding a
genotype able to modifying dynamically searching space and managing old
useful parts (Metalearning).
- Hi everyone,
I'm new to the NEAT group, so let me introduce myself shortly:
I live in Hungary and work as a software developer (currently at Lufthansa Systems), mostly in Java.
I'm interested in genetic/neural computing since university, but could play with it only in my free time.
I found NEAT some time ago and - as it seems to be a nice combination of the two fields - found it very interesting.
I've read the whole list (lots of brilliant ideas, very good value/noise rate, congratulations to everyone!), and collected about 50 interesting problems/ideas.
I decided to implement my own version of NEAT, mainly to have a better understanding of NEAT. Additionally I'd like a system which is as flexible as possible and reasonable, so I could integrate most of those nice ideas.
Using NEAT my main domain would be prediction of time series (e.g. stock prices). I've seen that there were some guys playing with this, but there were not too much (positive) results.
I wonder if those guys are already rich, or just their experiments didn't have too much succes...