Loading...
Please wait, while we are loading the content...
Similar Documents
Parallel Hybrid Meta-Heuristics: Application to the Quadratic Assignment Problem (1996)
| Content Provider | CiteSeerX |
|---|---|
| Author | Bachelet, V. Preux, P. Talbi, E. -G. |
| Description | : Meta-heuristics are search techniques that can be applied to a broad range of combinatorial optimization problems. Each meta-heuristic explores and exploits the search space in its own way. No heuristic can be better than any heuristic on a wide spectrum of problems. To make the search more efficient and robust, hybridization of heuristics should be used. In this paper, we present an ongoing research on parallel hybrid heuristics. The Quadratic Assignement Problem is used as a testbed problem. We present the performance of different meta-heuristics and their hybridization on standard problems taken from the QAPLIB library. Keywords: Parallel heuristic, Hybrid algorithm, Quadratic assignment problem, Hill-climbing, Genetic algorithm, Tabu search. 1 Introduction Unless P=NP, many interesting combinatorial optimization problems cannot be solved exactly within a reasonable amount of time. Consequently, heuristics must be used to solve large real-world problems. Heuristic algorithms may... |
| File Format | |
| Language | English |
| Publisher Date | 1996-01-01 |
| Publisher Institution | in Proceedings of the Parallel Optimization Colloquium |
| Access Restriction | Open |
| Subject Keyword | Wide Spectrum Genetic Algorithm Standard Problem Parallel Heuristic Search Technique Quadratic Assignement Problem Parallel Hybrid Heuristic Many Interesting Combinatorial Optimization Problem Search Space Reasonable Amount Qaplib Library Parallel Hybrid Meta-heuristics Different Meta-heuristics Ongoing Research Testbed Problem Tabu Search Combinatorial Optimization Problem Quadratic Assignment Problem Introduction Unless Np Hybrid Algorithm Heuristic Algorithm Broad Range Meta-heuristic Explores Large Real-world Problem |
| Content Type | Text |
| Resource Type | Article |