The influence of random number generators on graph partitioning algorithms
Electronic transactions on numerical analysis, Tome 21 (2005), pp. 125-133
Many modern heuristic algorithms rely at least in some part on random numbers. Using graph partitioning algorithms as an example, we demonstrate the often underestimated influence of these random-number generators on the result of the heuristic algorithms. Even methods that rely on random numbers mostly as a tiebreaking tool can deliver very different results depending only on the starting value of the pseudo-random number generator.
Classification :
05C85, 90C59, 94C15, 65Y99, 68R10
Keywords: graph partitioning, heuristic algorithms, pseudo random-number generator
Keywords: graph partitioning, heuristic algorithms, pseudo random-number generator
@article{ETNA_2005__21__a1,
author = {Elsner, Ulrich},
title = {The influence of random number generators on graph partitioning algorithms},
journal = {Electronic transactions on numerical analysis},
pages = {125--133},
year = {2005},
volume = {21},
zbl = {1112.68107},
language = {en},
url = {http://geodesic.mathdoc.fr/item/ETNA_2005__21__a1/}
}
Elsner, Ulrich. The influence of random number generators on graph partitioning algorithms. Electronic transactions on numerical analysis, Tome 21 (2005), pp. 125-133. http://geodesic.mathdoc.fr/item/ETNA_2005__21__a1/