Critical phenomena in a collective computation algorithm for combinatorial optimization problems
โ Scribed by T. Fukao; J. Wu; K. Ikeda
- Publisher
- Elsevier Science
- Year
- 1999
- Tongue
- English
- Weight
- 398 KB
- Volume
- 37
- Category
- Article
- ISSN
- 0898-1221
No coin nor oath required. For personal study only.
โฆ Synopsis
This paper discusses the critical temperature (control parameter) of an annealed neural network, a typical collective computation for combinatorial optimization problems.
It is shown that the theoretical critical temperatures determined by our estimation agree with those derived by computational experiments in graph partitioning problems and in the travelling salesman problem. (~) 1999 Elsevier Science Ltd. All rights reserved.
๐ SIMILAR VOLUMES
A genetic algorithm simulating Darwinian evolution is proposed to yield near-optimal solutions to the Traveling Salesman Problem. Noting that Darwinian evolution is itself an optimization process, we propose a heuristic algorithm that incorporates the tenets of natural selection. The time complexity