[Bbob-discuss] Request for BBOB algorithms.
olivier.teytaud at gmail.com
Tue Feb 28 10:07:14 CET 2012
> So, the questions are:****
> - Should the BBOB methodology be modified to really REQUIRE the info about
> the time demands of the algorithms?****
> - Should/can the COCO framework be modified to do this more automatically?
> - Should the time demands be more emhasized in the latex templates?
Thanks for comments; my personal answer would be yes for the last and
probably for all (even for "can", because I guess, if we do not try to
avoid people who cheat, that it is somehow feasible to measure computation
times) - I know it means that results of BBOB will take far more room on
PDF files, but I think it's necessary... just so that not only one side of
results is visible.
It's clear that computation times are unstable - depends on machine,
configuration, etc; but a factor of 20
between two (non parallel) machines is not so easy to get by such tricks,
so we could at least see what is
roughly feasible or not.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the bbob-discuss