Probability distribution of solution time in GRASP: An experimental investigation

Renata M. Aiex (rmainf.puc-rio.br)
Mauricio G. C. Resende (mgcrresearch.att.com)
Celso C. Ribeiro (celsoinf.puc-rio.br)

Abstract: A GRASP (greedy randomized adaptive search procedure) is a multi-start metaheuristic for combinatorial optimization. We study the probability distributions of solution time to a sub-optimal target value in five GRASPs that have appeared in the literature and for which source code is available. The distributions are estimated by running 12,000 independent runs of the heuristic. Standard methodology for graphical analysis is used to compare the empirical and theoretical distributions and estimate the parameters of the distributions. We conclude that the solution time to a sub-optimal target value fits a two-parameter exponential distribution. Hence, it is possible to approximately achieve linear speed-up by implementing GRASP in parallel.

Keywords: Combinatorial optimization, meta-heuristic, GRASP, experimental analysis of algorithms, probability distribution, parallel algorithm

Category 1: Combinatorial Optimization (Meta Heuristics )

Citation: AT&T Labs Research Technical Report, October 9, 2000

Download: [PDF]

Entry Submitted: 02/16/2001
Entry Accepted: 02/16/2001
Entry Last Modified: 02/19/2001

Modify/Update this entry