In this and two companion papers, we report on an extended empirical study of the simulated annealing approach to combinatorial optimization proposed by S. Kirkpatrick et al. That study investigated how best to adapt simulated annealing to particular problems and compared its performance to that of more traditional algorithms. This paper (Part I) discusses annealing and our parameterized generic implementation of it, describes how we adapted this generic algorithm to the graph partitioning problem, and reports how well it compared to standard algorithms like the Kernighan-Lin algorithm. (For sparse random graphs, it tended to outperform Kernighan-Lin as the number of vertices become large, even when its much greater running time was taken into account. It did not perform nearly so well, however, on graphs generated with a built-in geometric structure.) We also discuss how we went about optimizing our implementation, and describe the effects of changing the various annealing parameters or varying the basic annealing algorithm itself.
This is the second in a series of three papers that empirically examine the competitiveness of simulated annealing in certain well-studied domains of combinatorial optimization. Simulated annealing is a randomized technique proposed by S. Kirkpatrick, C. D. Gelatt and M. P. Vecchi for improving local optimization algorithms. Here we report on experiments at adapting simulated annealing to graph coloring and number partitioning, two problems for which local optimization had not previously been thought suitable. For graph coloring, we report on three simulated annealing schemes, all of which can dominate traditional techniques for certain types of graphs, at least when large amounts of computing time are available. For number partitioning, simulated annealing is not competitive with the differencing algorithm of N. Karmarkar and R. M. Karp, except on relatively small instances. Moreover, if running time is taken into account, natural annealing schemes cannot even outperform multiple random runs of the local optimization algorithms on which they are based, in sharp contrast to the observed performance of annealing on other problems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.