Problem partitioning in hybrid genetic algorithms

  • Authors:

  • Philip Little;Bart Rylander

  • Affiliations:

  • School of Engineering, University of Portland, Portland;School of Engineering, University of Portland, Portland

  • Venue:
  • CSECS'06 Proceedings of the 5th WSEAS International Conference on Circuits, Systems, Electronics, Control & Signal Processing
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Genetic algorithms are well suited to hybridization. Problem division between components of the hybrid varies by problem type and by the mechanisms included in the hybrid. We propose a hybridization technique for combining genetic algorithms (GAs) and deterministic algorithms based on solution candidate partitioning. We conduct a set of experiments to evaluate several instances of this hybridization scheme and demonstrate the efficiency of this hybridization within certain partition ranges for these problems. These results indicate that performance of this hybrid approach is superior to a GA or a deterministic algorithm alone for the problem instances examined, a result that may hold for problems of the same classes as those examined. Finally, we propose topics for future research.