Go to Google Groups Home    Computational Intelligence and Co-evolution
Re: [coevolve] CEC report

Julian Togelius <julian.togel...@gmail.com>

Hi everyone, I have my own impressions from CEC at:

http://togelius.blogspot.com/2007/10/cec-2007-conference-report.html

A special session on co-evolution is certainly a good idea - I would
definitely try to attend, whether I would submit anything depends on
what directions my research will take after I start at IDSIA.

Julian

On 10/10/07, phi <PhilipHings...@gmail.com> wrote:

> Hi everyone

> I thought I would post a brief report on the recent CEC in Singapore.
> It's brief because, unfortunately, I got a stomach bug early on in the
> conference and had to spend the rest of the time close to my bathroom.
> So I didn't get to many talks and missed out on catching up with lots
> of people. If anyone else was there, maybe you could fill in some
> gaps! Evan? Julian? Enyone?

> I did manage to get to a couple of the keynote talks.

> One was by Xin Yao: "A Rigorous Theoretical Framework for Measuring
> Generalisation of Co-evolutionary Learning" . See
> http://cec2007.nus.edu.sg/staticPages/plenary%20speakers.html#ARigorous
> . This was a very good talk with some interesting new ideas.

> Simon Lucas gave a talk about CI in games: "Games, and The Design of
> New Hybrid Evolutionary and Temporal Difference Learning Algorithms".
> Many of his examples used co-evolution. Actually, I chaired a special
> session on CI in games. Of the 6 talks, 5 were about co-evolution
> (mine was the only one that wasn't!) See
> http://cec2007.nus.edu.sg/staticPages/plenary%20speakers.html#GamesAnd
> for an abstract of Simon's talk.

> Incidentally, no-one has so far responded to the suggestion of running
> a Symposium on Co-evolution at the next SSCI - fair enough, it would
> be a major exercise. It would be a good way to broaden the audience
> for co-evolutionary theory though - there was a lot of interchange of
> people between the different symposia at SSCI in Hawaii - more than
> at, say, a WCCI.

> But how about a special session on co-evolution at WCCI in Hong Kong
> next year? Any takers for that? The workload for a special session is
> not that high and it is quite straightforward to run one.

> regards, phi

--
Julian Togelius
Department of Computing and Electronic Systems
University of Essex
Colchester, CO4 3SQ, UK
jul...@togelius.com
http://julian.togelius.com
+44-7851-742111
+46-705-192088