Innovative strategy of SOMA control parameter setting

  • Authors:
  • Pavel Vařacha

  • Affiliations:
  • Tomas Bata University in Zlin, Faculty of Applied Informatics, Zlin, Czech Republic

  • Venue:
  • NNECFSIC'12 Proceedings of the 12th WSEAS international conference on Neural networks, fuzzy systems, evolutionary computing & automation
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

SOMA - Self-Organizing Migration Algorithm is highly effective tool of evolutionary optimization as already proven on many real life problems. This article introduces new innovative strategy how to set one of the most important SOMA control parameter PRT, which makes the algorithm even more efficient. 996.3E6 single evaluations have been calculated for 10 different test functions to statistically prove impact of this improvement. Application of the strategy proposed by this article can bring better results in about 50% of optimized functions and possibly even achieve breakthrough in cases, which standard SOMA setting does not perform well.