Name
Abstract | ||
|---|---|---|
Parallelization is becoming a more important issue for solving difficult optimization problems. Island models combine phases of independent evolution with migration where genetic information is spread out to neighbored islands. This can lead to an increased diversity within the population. Despite many successful applications, the reasons behind the success of island models are not well understood. We perform a first rigorous runtime analysis for island models and construct a function where phases of independent evolution as well as communication among the islands are essential. A simple island model with migration finds a global optimum in polynomial time, while panmictic populations as well as island models without migration need exponential time, with very high probability. Our results lead to new insights into the usefulness of migration, how information is propagated in island models, and how to set parameters such as the migration interval. This is a novel contribution to the theoretical foundation of parallel EAs. Further, we provide empirical results that complement the theoretical results, investigate the robustness with respect to the choice of the migration interval and compare various migration topologies using statistical tests. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1007/s00500-013-0991-0 | Soft Computing - A Fusion of Foundations, Methodologies and Applications |
Keywords | Field | DocType |
runtime analysis,spatial structures,parallel evolutionary algorithms,multi-deme model,distributed evolutionary algorithms,migration,island model | Population,Mathematical optimization,Exponential function,Evolutionary algorithm,Computer science,Theoretical computer science,Robustness (computer science),Network topology,Time complexity,Optimization problem,Statistical hypothesis testing | Journal |
Volume | Issue | ISSN |
17 | 7 | 1433-7479 |
Citations | PageRank | References |
17 | 0.63 | 25 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jörg Lässig | 1 | 175 | 22.53 |
| Dirk Sudholt | 2 | 1063 | 64.62 |