 | CATS |  |
Experimental Methodology
The raison d'être of CATS is experimental research in
discrete and combinatorial algorithms. Yet the area is relatively new
and methodology to date is largely ad hoc. The major contribution
of CATS to methodology is its enabling of comparisons between various
solution methods. However, what should be measured and how remain
difficult questions that each researcher should ponder carefully before
embarking on a series of experiments. Guidance, or at least examples,
can be found in the articles published in the
ACM Journal of Experimental Algorithmics
as well as in a forthcoming collection of papers edited by David Johnson
in the DIMACS Monograph series, resulting from an
Experimental Methodology
Day held as part of a DIMACS Challenge Workshop; included is
a review of the field by the editor-in-chief
of the ACM JEA.