MOCell
Description
MOCell (MultiObjective Cellular genetic algorithm) is a family of cellular genetic algorithms (cGAs) for multi-objective optimization which are the result of combining different strategies (synchronous vs asynchronous and two archive feedback schemes).
Main features
- Adaptation of the canonical cGA scheme to deal with multi-objective problems
-
External archive to store the non-dominated solutions
- Density estimator: crowding distance
- Feedback: the solutions of the archive are used to enhance the search
- Neighbohood: 1 hop neighbours (8 surrounding solutions)
- Operators: SBX crossover and polynomial mutation
Configurations
Six configurations of MOCell have been proposed:
- sMOCell1: the original synchronous MOCell algorithm
- sMOCell2: MOCell + archive feedback through parent selection
- aMOCell1: asynchronous MOCell
- aMOCell2: aMOCell1 + archive feedback through parent selection
- aMOCell3: aMOCell1 + replacing the worst neighbor
- aMOCell4: Combination of aMOCell2 and aMOCell3
In PPSN 2009 we presented a variant of aMOCell3, named CellDE, which replaced the SBX crossover and polynomial mutation by the differential evolution operators
MOCell steps
References
- A.J. Nebro, J.J. Durillo, F. Luna, B. Dorronsoro, E. Alba. MOCell: A Cellular Genetic Algorithm for Multiobjective Optimization. International Journal of Intelligent Systems. Vol.24, No. 7 (July 2009), pp. 726-746. DOI BIBTEX
- A.J. Nebro, J.J. Durillo, F. Luna, B. Dorronsoro, E. Alba Design Issues in a Multiobjective Cellular Genetic Algorithm. Evolutionary Multi-Criterion Optimization. 4th International Conference, EMO 2007. Sendai/Matsushima, Japan, March 2007. BIBTEX
- J.J. Durillo, A.J. Nebro, F. Luna, E. Alba Solving Three-Objective Optimization Problems Using a new Hybrid Cellular Genetic Algorithm. PPSN X. LNCS, Vol. Volume 5199/2008, pp: 661-670. Dortmund, September 2008. DOI BIBTEX