 |
 |
 |
 |
|
|
|
|
These competition are now closed. See competition results for list of winners and more details
The following competitions were run for WCCI 2002. This page contains a brief summary of each problem, together with links to more details of what the problem involves and how to enter a solution. See the individual competition pages for the important dates as they differ for each one. Please contact the competition organizer if you have any queries or problems.
In order to be eligible for any prize you must be registered for WCCI 2002.
There is a $100 prize for the winner of each competition (courtesy of the Evolutionary Programming Society), plus complimentary registration for IJCNN 2003, FUZZ-IEEE 2003 or CEC2003 at the winners' discretion.
Also, soda, the people who make sodaplay, are giving away sodarace tee-shirts to the first 80 entrants. So, don't delay, enter today! (note: tee-shirts must be collected at the conference).
| Competition | Organizer | Contact |
|---|---|---|
| 3D Design Optimization | Bernhard Sendhoff | mailto:bs@el-tec.de |
| Checkers | Simon Lucas | mailto:sml@essex.ac.uk |
| Symbol sequence recognition | Simon Lucas | mailto:sml@essex.ac.uk |
| Time Series Prediction | Simon Lucas | mailto:sml@essex.ac.uk |
In the 3D design optimization task, an unknown three-dimensional surface has
to be approximated as closely as possible. The surface is closed in both
parameter directions and describes a 3D structure like a bottle or a
teapot. The target surface is smooth, has no
holes, i.e., it is not like a donut, and does not intersect
itself. Thus, it can be represented by a closed and periodic
Non-Uniform Rational B-Spline (NURBS) surface.
More details
The problem setup comes with all the code you need to participate in a
computer Checkers competition over the Web. You run the algorithms of your
choice on your own machine, and connect to a game server via the supplied Java
client. The supplied Java client plays a game against a remote
opponent by exchanging XML messages with the server.
More
details
The object of this competition is to infer a symbol sequence recognizer from a set of training data. The problem is specified in general terms, and may include datasets from problems in bioinformatics, optical character recognition and language modeling. Note that your algorithm sees a standard interface to any of the datasets, and will not told about the nature of each dataset.
Predict the future by observing the past.