Polymorphic Random Building Block Operator for Genetic Algorithms

Ghodrat Moghadampour

2012

Abstract

Boosting the evolutionary process of genetic algorithms by generating better individuals, avoiding stagnation at local optima and refreshing population in a desirable way is a challenging task. Typically operators are used to achieve these objectives. On the other hand using operators can become a challenging task in itself if applying them requires setting many parameters through human intervention. Therefore, developing operators, which do not require human intervention and at the same time are capable of assisting the evolutionary process, is highly desirable. Most typical genetic operators are mutation and crossover. However, experience has proved that these operators in their classical form are not capable of refining the population efficiently enough. In this work a new dynamic mutation operator called polymorphic random building block operator with variable mutation rate is proposed. This operator does not require any pre-fixed parameter. It randomly selects a section from the binary presentation of the individual, then generates a random bit-string of the same length as the selected section and applies bitwise logical AND, OR and XOR operators between the randomly generated bit-string and the selected section from the individual. In the next step all three newly generated offspring will go through selection procedure and will replace a possibly worse individual in the population. Experimentation with 33 test functions and 11550 test runs proved the superiority of the proposed dynamic mutation operator over single-point mutation operator with 1%, 5% and 8% mutation rates and the multipoint mutation operator with 5%, 8% and 15% mutation rates.

References

  1. Bäck, Thomas, David B. Fogel, Darrell Whitely & Peter J. Angeline, 2000. Mutation operators. In: Evolutionary Computation 1, Basic Algorithms and Operators. Eds T. Bäck, D.B. Fogel & Z. Michalewicz. United Kingdom: Institute of Physics Publishing Ltd, Bristol and Philadelphia. ISBN 0750306645.
  2. De Jong, K. A., 1975. An Analysis of the Behavior of a Class of Genetic Adaptive Systems. Ph.D. thesis, University of Michigan. Michigan: Ann Arbor.
  3. Eiben, A. and J. Smith, 2007. Introduction to Evolutionary Computing. Natural Computing Series. Springer, 2nd edition.
  4. Eiben, G. and M. C. Schut, 2008. New Ways To Calibrate Evolutionary Algorithms. In Advances in Metaheuristics for Hard Optimization, pages 153-177.
  5. Eshelman, L. J. & J. D. Schaffer, 1991. Preventing premature convergence in genetic algorithms by preventing incest. In Proceedings of the Fourth International Conference on Genetic Algorithms. Eds. R. K. Belew & L. B. Booker. San Mateo, CA : Morgan Kaufmann Publishers.
  6. Holland, J. H., 1975. Adaptation in Natural and Artificial Systems. Ann Arbor: MI: University of Michigan Press.
  7. Mengshoel, Ole J. & Goldberg, David E., 2008. The crowding approach to niching in genetic algorithms. Evolutionary Computation, Volume 16 , Issue 3 (Fall 2008). ISSN:1063-6560.
  8. Michalewicz, Zbigniew (1996). Genetic Algorithms + Data Structures = Evolution Programs. Third, Revised and Extended Edition. USA: Springer. ISBN 3-540-60676-9.
  9. Michalewicz, Zbigniew, 2000. Introduction to search operators. In Evolutionary Computation 1, Basic Algorithms and Operators. Eds T. Bäck, D.B. Fogel & Z. Michalewicz. United Kingdom: Institute of Physics Publishing Ltd, Bristol and Philadelphia. ISBN 0750306645.
  10. Mitchell, Melanie, 1998. An Introducton to Genetic Algorithms. United States of America: A Bradford Book. First MIT Press Paperback Edition.
  11. Moghadampour, Ghodrat (2006). Genetic Algorithms, Parameter Control and Function Optimization: A New Approach. PhD dissertation. ACTA WASAENSIA 160, Vaasa, Finland. ISBN 952-476-140-8.
  12. Moghadampour, Ghodrat (2011). Random Building Block Operator for Genetic Algorithms. 13th International Conference on Enterprise Information Systems (ICEIS 2011), 08 - 11 June 2011Beijing - China.
  13. Moghadampour, Ghodrat (2012). Outperforming Mutation Operator with Random Building Block Operator in Genetic Algorithms. In Enterprise Information Systems International Conference, ICEIS 2011 Beijing, China, June 8-11, 2011 Revised Selected Papers. Eds. Runtong Zhang, Zhenji Zhang, Juliang Zhang, Joaquim Filipe and José Cordeiro. SpringerVerlag LNBIP Series book.
  14. Mühlenbein, H., 1992. How genetic algorithms really work: 1. mutation and hill-climbing. In: Parallel Problem Solving from Nature 2. Eds R. Männer & B. Manderick. North-Holland.
  15. Smit, S. K. and Eiben, A. E., 2009. Comparing Parameter Tuning Methods for Evolutionary Algorithms. In IEEE Congress on Evolutionary Computation (CEC), pages 399-406, May 2009.
  16. Spears, W. M., 1993. Crossover or mutation? In: Foundations of Genetic Algorithms 2. Ed. L. D. Whitely. Morgan Kaufmann.
  17. Ursem, Rasmus K., 2003. Models for Evolutionary Algorithms and Their Applications in System Identification and Control Optimization (PhD Dissertation). A Dissertation Presented to the Faculty of Science of the University of Aarhus in Partial Fulfillment of the Requirements for the PhD Degree. Department of Computer Science, University of Aarhus, Denmark.
  18. Whitley, Darrell, 2000. Permutations. In Evolutionary Computation 1, Basic Algorithms and Operators. Eds T. Bäck, D.B. Fogel & Z. Michalewicz. United Kingdom: Institute of Physics Publishing Ltd, Bristol and Philadelphia. ISBN 0750306645.
Download


Paper Citation


in Harvard Style

Moghadampour G. (2012). Polymorphic Random Building Block Operator for Genetic Algorithms . In Proceedings of the 14th International Conference on Enterprise Information Systems - Volume 1: ICEIS, ISBN 978-989-8565-10-5, pages 342-348. DOI: 10.5220/0004014103420348


in Bibtex Style

@conference{iceis12,
author={Ghodrat Moghadampour},
title={Polymorphic Random Building Block Operator for Genetic Algorithms},
booktitle={Proceedings of the 14th International Conference on Enterprise Information Systems - Volume 1: ICEIS,},
year={2012},
pages={342-348},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004014103420348},
isbn={978-989-8565-10-5},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 14th International Conference on Enterprise Information Systems - Volume 1: ICEIS,
TI - Polymorphic Random Building Block Operator for Genetic Algorithms
SN - 978-989-8565-10-5
AU - Moghadampour G.
PY - 2012
SP - 342
EP - 348
DO - 10.5220/0004014103420348