Computational Optimization and Applications

, Volume 56, Issue 1, pp 209–229

Parameter-less algorithm for evolutionary-based optimization

For continuous and combinatorial problems
Article

DOI: 10.1007/s10589-013-9565-4

Cite this article as:
Papa, G. Comput Optim Appl (2013) 56: 209. doi:10.1007/s10589-013-9565-4

Abstract

The development of a simple, adaptive, parameter-less search algorithm was initiated by the need for an algorithm that is able to find optimal solutions relatively quick, and without the need for a control-parameter-setting specialist. Its control parameters are calculated during the optimization process, according to the progress of the search. The algorithm is intended for continuous and combinatorial problems. The efficiency of the proposed parameter-less algorithm was evaluated using one theoretical and three real-world industrial optimization problems. A comparison with other evolutionary approaches shows that the presented adaptive parameter-less algorithm has a competitive convergence with regards to the comparable algorithms. Also, it proves algorithm’s ability to finding the optimal solutions without the need for predefined control parameters.

Keywords

Optimization Evolutionary Adaptive Parameter-less Continuous Combinatorial 

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  1. 1.Jožef Stefan InstituteLjubljanaSlovenia