Virtual loser genetic algorithm for dynamic environments

Authors:
Anabela Simões;Ernesto Costa
Affiliations:
Coimbra Institute of Engineering, Polytechnic Institute of Coimbra, Portugal and Centre for Informatics and Systems of the University of Coimbra, Portugal;Centre for Informatics and Systems of the University of Coimbra, Portugal
Venue:
EvoApplications'12 Proceedings of the 2012t European conference on Applications of Evolutionary Computation
Year:
2012

Citing 6
Cited 2

Evolutionary Optimization in Dynamic Environments

Evolutionary Optimization in Dynamic Environments
Prediction in evolutionary algorithms for dynamic environments using markov chains and nonlinear regression

Proceedings of the 11th Annual conference on Genetic and evolutionary computation
Variable-Size Memory Evolutionary Algorithm to Deal with Dynamic Environments

Proceedings of the 2007 EvoWorkshops 2007 on EvoCoMnet, EvoFIN, EvoIASP,EvoINTERACTION, EvoMUSART, EvoSTOC and EvoTransLog: Applications of Evolutionary Computing
Hyper-learning for population-based incremental learning in dynamic environments

CEC'09 Proceedings of the Eleventh conference on Congress on Evolutionary Computation
A memory enhanced evolutionary algorithm for dynamic scheduling problems

Evo'08 Proceedings of the 2008 conference on Applications of evolutionary computing
Population-Based Incremental Learning With Associative Memory for Dynamic Environments

IEEE Transactions on Evolutionary Computation

Enhancing the virtual loser genetic algorithm for dynamic environments

Proceedings of the 14th annual conference on Genetic and evolutionary computation
Extended virtual loser genetic algorithm for the dynamic traveling salesman problem

Proceedings of the 15th annual conference on Genetic and evolutionary computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Memory-based Evolutionary Algorithms in Dynamic Optimization Problems (DOPs) store the best solutions in order to reuse them in future situations. The memorization of the best solutions can be direct (the best individual of the current population is stored) or associative (additional information from the current population is also stored). This paper explores a different type of associative memory to use in Evolutionary Algorithms for DOPs. The memory stores the current best individual and a vector of inhibitions that reflect past errors performed during the evolutionary process. When a change is detected in the environment the best solution is retrieved from memory and the vector of inhibitions associated to this individual is used to create new solutions avoiding the repetition of past errors. This algorithm is called Virtual Loser Genetic Algorithm and was tested in different dynamic environments created using the XOR DOP generator. The results show that the proposed memory scheme significantly enhances the Evolutionary Algorithms in cyclic dynamic environments.