ABSTRACT
In this work, we explore the idea that parameter setting of stochastic metaheuristics should be considered as a multi-objective problem. The so-called "performance fronts" presented in this work are a collection of non-dominated parameters sets, satisfying both a speed and a precision objective. Experiments are conducted using a multi-objective evolutionary algorithm, in order to: (i) set a parameter of several continuous metaheuristics, and (ii) set parameters of an hybrid algorithm for temporal planning.
Our results suggest that the performance fronts are well suited for setting the parameters of stochastic metaheuristics. The relative position, in the objective space, of several parameter fronts also permits to compare metaheuristics on a given problem. Moreover, this approach give insights on the algorithm behaviour.
- ]]T. Bartz-Beielstein. Experimental Research in Evolutionary Computation: The New Experimentalism. Natural Computing Series. Springer, 2006. Google ScholarDigital Library
- ]]T. Bartz-Beielstein, C. W. G. Lasarczyk, and M. Preuss. Sequential parameter optimization. volume 1, pages 773--780, 2005.Google Scholar
- ]]M. Birattari, T. Stutzle, L. Paquete, and K. Varrentrapp. A racing algorithm for configuring metaheuristics. In GECCO '02: Proceedings of the Genetic and Evolutionary Computation Conference, pages 11--18, San Francisco, CA, USA, 2002. Morgan Kaufmann Publishers Inc.Google ScholarDigital Library
- ]]S. Cahon, N. Melab, and E. G. Talbi. Paradiseo: A framework for the reusable design of parallel and distributed metaheuristics. Journal of Heuristics, 10(3):357--380, 2004. Google ScholarDigital Library
- ]]J. Clune, S. Goings, B. Punch, and E. Goodman. Investigations in meta-gas: panaceas or pipe dreams? In GECCO '05: Proceedings of the 2005 workshops on Genetic and evolutionary computation, pages 235--241, New York, NY, USA, 2005. ACM. Google ScholarDigital Library
- ]]J. Dreo. Multi-criteria meta-parameter tuning for mono-objective stochastic metaheuristics. In 2nd International Conference on Metaheuristics and Nature Inspired Computing, October 2008.Google Scholar
- ]]J. Dreo, J.-P. Aumasson, W. Tfaili, and P. Siarry. Adaptive learning search, a new tool to help comprehending metaheuristics. International Journal on Artificial Intelligence Tools, 16(3), june 2007.Google Scholar
- ]]J. Dreo and P. Siarry. Stochastic metaheuristics as sampling techniques using swarm intelligence. I-Tech Education and Publishing, December 2007.Google ScholarCross Ref
- ]]F. G. Lobo, C. F. Lima, and Z. Michalewicz, editors. Parameter Setting in Evolutionary Algorithms, volume 54 of Studies in Computational Intelligence. Springer, 2007. Google ScholarDigital Library
- ]]D. Long and M. Fox. The 3rd international planning competition: Results and analysis. Journal of Artificial Intelligence Research, 20:1--59, 2003. Google ScholarCross Ref
- ]]V. Nannen, S. K. Smit, and A. E. Eiben. Costs and benefits of tuning parameters of evolutionary algorithms. In Proceedings of the 10th international conference on Parallel Problem Solving from Nature, pages 528--538, Berlin, Heidelberg, 2008. Springer-Verlag.Google ScholarCross Ref
- ]]M. Schoenauer, P. Saveant, and V. Vidal. Divide-and-evolve: A new memetic scheme for domain-independent temporal planning. pages 247--260. 2006. Google ScholarDigital Library
- ]]H.-P. P. Schwefel. Evolution and Optimum Seeking: The Sixth Generation. John Wiley & Sons, Inc., New York, NY, USA, 1993. Google ScholarDigital Library
Index Terms
- Using performance fronts for parameter setting of stochastic metaheuristics
Recommendations
Optimization of assignment of tasks to teams using multi-objective metaheuristics
GECCO '13 Companion: Proceedings of the 15th annual conference companion on Genetic and evolutionary computationA highly interesting but not thoroughly addressed optimization problem is a variation of the Assignment Problem (AP) where tasks are assigned to groups of collaborating agents (teams). In this paper, we address this class of AP as a bi-objective ...
Structural design using multi-objective metaheuristics. Comparative study and application to a real-world problem
Many structural design problems in the field of civil engineering are naturally multi-criteria, i.e., they have several conflicting objectives that have to be optimized simultaneously. An example is when we aim to reduce the weight of a structure while ...
Improving the performance of MO-RV-GOMEA on problems with many objectives using tchebycheff scalarizations
GECCO '18: Proceedings of the Genetic and Evolutionary Computation ConferenceThe Multi-Objective Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (MO-RV-GOMEA) has been shown to exhibit excellent performance in solving various bi-objective benchmark and real-world problems. We assess the competence of MO-RV-GOMEA in ...
Comments