SETI@home: an experiment in public-resource computing
Communications of the ACM
A Framework for Distributed Evolutionary Algorithms
PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Spatially Structured Evolutionary Algorithms: Artificial Evolution in Space and Time (Natural Computing Series)
Gossip-based aggregation in large dynamic networks
ACM Transactions on Computer Systems (TOCS)
Exploring selection mechanisms for an agent-based distributed evolutionary algorithm
Proceedings of the 9th annual conference companion on Genetic and evolutionary computation
Peer-to-peer evolutionary algorithms with adaptive autonomous selection
Proceedings of the 9th annual conference on Genetic and evolutionary computation
Autonomous selection in evolutionary algorithms
Proceedings of the 9th annual conference on Genetic and evolutionary computation
Effects of scale-free and small-world topologies on binary coded self-adaptive CEA
EvoCOP'06 Proceedings of the 6th European conference on Evolutionary Computation in Combinatorial Optimization
Characterizing fault-tolerance of genetic algorithms in desktop grid systems
EvoCOP'10 Proceedings of the 10th European conference on Evolutionary Computation in Combinatorial Optimization
Hi-index | 0.00 |
In this paper we propose an improvement on a fully distributed Peer-to-Peer (P2P) Evolutionary Algorithm (EA) based on autonomous selection. Autonomous selection means that individuals decide on their own state of reproduction and survival without any central control, using instead estimations about the global population state for decision making. The population size varies at run-time as a consequence of such a decentralized reproduction and death of individuals. In order to keep it stable, we propose a self-adjusting mechanism which has been shown successful in three different search landscapes. Key are the estimations about fitness and size of the population as provided by a gossiping algorithm. Such an algorithm requires several rounds to collect the information while the individuals have to wait for synchronization. As an improvement, we propose a completely asynchronous EA which does not need waiting times. The results show that our approach outperforms quantitatively the execution time of the synchronous version.