A New Stopping Criterion for Genetic Algorithms

Christelle Reynes, Robert Sabatier

2012

Abstract

Obtaining theoretically legitimate stopping criteria is a difficult task. Being able to use such criteria, especially in real-encoding context, remains an open problem. The proposed criterion is based on a Markov chain modelling and on the distribution of the number of occurrences of the locally best solution during several generations under the assumption of non-convergence. The algorithm stops when the probability of obtaining the observed number of occurrences is too small. The obtained criterion is able to fit very different solution spaces and fitness functions (within studied limitations) without any required user intervention.

References

  1. Aytug, H. and Koehler, G. (2000). New stopping criterion for genetic algorithms. European Journal of Operational Research, 126(3):662-674.
  2. Davis, T. and Principe, J. (1993). A markov chain framework for the simple genetic algorithm. Evolutionary computation, 1(3):269-288.
  3. Moré, J., Garbow, B., and Hillstrom, K. (1981). Testing unconstrained optimization software. ACM Transactions on Mathematical Software (TOMS), 7(1):17-41.
  4. Mühlenbein, H., Schomisch, M., and Born, J. (1991). The parallel genetic algorithm as function optimizer. Parallel computing, 17(6-7):619-632.
  5. Storch, T. (2008). On the choice of the parent population size*. Evolutionary Computation, 16(4):557-578.
Download


Paper Citation


in Harvard Style

Reynes C. and Sabatier R. (2012). A New Stopping Criterion for Genetic Algorithms . In Proceedings of the 4th International Joint Conference on Computational Intelligence - Volume 1: ECTA, (IJCCI 2012) ISBN 978-989-8565-33-4, pages 202-207. DOI: 10.5220/0004154202020207


in Bibtex Style

@conference{ecta12,
author={Christelle Reynes and Robert Sabatier},
title={A New Stopping Criterion for Genetic Algorithms},
booktitle={Proceedings of the 4th International Joint Conference on Computational Intelligence - Volume 1: ECTA, (IJCCI 2012)},
year={2012},
pages={202-207},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004154202020207},
isbn={978-989-8565-33-4},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 4th International Joint Conference on Computational Intelligence - Volume 1: ECTA, (IJCCI 2012)
TI - A New Stopping Criterion for Genetic Algorithms
SN - 978-989-8565-33-4
AU - Reynes C.
AU - Sabatier R.
PY - 2012
SP - 202
EP - 207
DO - 10.5220/0004154202020207