Advertisement

Inside GA dynamics: Ground basis for comparison

  • Leila Kallel
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1498)

Abstract

Much attention has been paid in the GA literature to understand, characterize or to predict the notion of difficulty for genetic algorithms. Formal or informal ways of handling difficulty in previous work are commented. This points out a major problem of scaling especially when comparing differently distributed fitness functions.

Hamming fitness functions are proposed as a basis to scale difficulty measures, and to account for GA parameters bias. The use of a basis relaxes the dependence on fitness scale or distribution. Different measures are also proposed to characterize GA behavior, distinguishing convergence time and on-line GA radial and effective trajectory distance in Hamming space.

Keywords

Genetic Algorithm Fitness Function Convergence Time Fitness Landscape Problem Difficulty 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    S. Baluja. An empirical comparison of seven iterative and evolutionary function optimization heuristics. Technical Report CMU-CS-95-193, Carnegie Mellon University, 1995.Google Scholar
  2. 2.
    Robert J. Collins and David R. Jefferson. Selection in massively parallel genetic algorithm. In R. K. Belew and L. B. Booker, editors, Proc. of the 4th ICGA, pages 249–256. Morgan Kaufmann, 1991.Google Scholar
  3. 3.
    J. Culberson and Jonathan Lichtner. On searching ‘alpha'-ary hypercubes and related graphs. In R. K. Belew and M. D. Vose, editors, FOGA 4, pages 263–290. Morgan Kaufmann, 1996.Google Scholar
  4. 4.
    K. Mathias D. Whitley and L. Pyeatt. Hyperplane ranking in simple genetic alghoritms. In L. J. Eshelman, editor, Proc. of the 6th ICGA, pages 231–238. Morgan Kaufmann, 1995.Google Scholar
  5. 5.
    Y. Davidor. Epistasis variance: a viewpoint on ga-hardness. In G. J. E. Rawlins, editor, FOGA, pages 23–35. Morgan Kaufmann, 1991.Google Scholar
  6. 6.
    S. Forrest and M. Mitchell. The performance of genetic algorithms on walsh polynomials: Some anomalous results and their explanation. In R. K. Belew and L. B. Booker, editors, Proc. of the 4th ICGA, pages 182–189. Morgan Kaufmann, 1991.Google Scholar
  7. 7.
    S. Forrest and M. Mitchell. Relative building-block fitness and the building-block hypothesis. FOGA 2, pages 109–126, 1993.Google Scholar
  8. 8.
    D. E. Goldberg. Genetic algorithms in search, optimization and machine learning. Addison Wesley, 1989.Google Scholar
  9. 9.
    D. E. Goldberg. Making genetic algorithms fly: a lesson from the wright brothers. Advanced Technology for Developpers, 2:1–8, February 1993.zbMATHGoogle Scholar
  10. 10.
    J. Horn and D.E. Goldberg. Genetic algorithms difficulty and the modality of fitness landscapes. In L. D. Whitley and M. D. Vose, editors, FOGA 3, pages 243–269. Morgan Kaufmann, 1995.Google Scholar
  11. 11.
    J. Horn, D.E. Goldberg, and K. Deb. Long path problems. In Y. Davidor, H.-P. Schwefel, and R. Manner, editors, Proc. of the 3rd Conference on Parallel Problems Solving from Nature, pages 149–158. Springer Verlag, 1994.Google Scholar
  12. 12.
    T. Jones and S. Forrest. Fitness distance correlation as a measure of problem difficulty for genetic algorithms. In L. J. Eshelman, editor, Proc. of the 6th ICGA, pages 184–192. Morgan Kaufmann, 1995.Google Scholar
  13. 13.
    L. Kallel and M. Schoenauer. Alternative random initialization in genetic algorithms. In Th. Bäeck, editor, Proc. of the 7th ICGA. Morgan Kaufmann, 1997.Google Scholar
  14. 14.
    H. Kargupta and D. E. Goldberg. Serch, blackbox optimisation, and sample complexity. In R. K. Belew and M. D. Vose, editors, FOGA 4, pages 291–324. Morgan Kaufmann, 1996.Google Scholar
  15. 15.
    S. A. Kauffman. Lectures in the Sciences of Complexity, volume I of SFI studies, chapter Adaptation on Rugged Fitness Landscape, pages 619–712. Addison Wesley, 1989.Google Scholar
  16. 16.
    B. Manderick, M. de Weger, and P. Spiessens. The genetic algorithm and the structure of the fitness landscape. In R. K. Belew and L. B. Booker, editors, Proc. of the 4th ICGA, pages 143–150. Morgan Kaufmann, 1991.Google Scholar
  17. 17.
    M. Mitchell and J.H. Holland. When will a genetic algorithm outperform hillclimbing ? In S. Forrest, editor, Proc. of the 5th ICGA, page 647, 1993.Google Scholar
  18. 18.
    B. Naudts and L.Kallel. Some facts about so called ga-hardness measures. Technical Report 379, CMAP, Ecole Polytechnique, Mars 1998. submitted.Google Scholar
  19. 19.
    N. J. Radcliffe and P. D. Surry. Fitness variance of formae and performance prediction. In L. D. Whitley and M. D. Vose, editors, FOGA 3, pages 51–72. Morgan Kaufmann, 1995.Google Scholar
  20. 20.
    S. Rochet, G. Venturini, M. Slimane, and E.M El Kharoubi. A critical and empirical study of epistasis measures for predicting ga performances: a summary. In J.-K. Hao, E. Lutton, E. Ronald, M. Schoenauer, and D. Snyers, editors, Artificial Evolution'97, LNCS, pages 287–299. Springer Verlag, 1997.Google Scholar
  21. 21.
    J. D. Schaffer, R. A. Caruana, L. Eshelman, and R. Das. A study of control parameters affecting on-line performance of genetic algorithms for function optimization. In J. D. Schaffer, editor, Proc. of the 3rd ICGA, pages 51–60. Morgan Kaufmann, 1989.Google Scholar
  22. 22.
    K. A. DeJong W. M. Spears and D. M. Gordon. Using markov chains to analyse gafos. In L. D. Whitley and M. D. Vose, editors, FOGA 3, pages 115–137. Morgan Kaufmann, 1995.Google Scholar
  23. 23.
    Michael D. Vose and Alden H. Wright. Stability of vertex fixed points and applications. In L. D. Whitley and M. D. Vose, editors, FOGA 3, pages 103–11. Morgan Kaufmann, 1995.Google Scholar
  24. 24.
    D. Whitley. Fundamental principles of deception in genetic search. In G. J. E. Rawlins, editor, FOGA. Morgan Kaufmann, 1991.Google Scholar
  25. 25.
    S.W. Wilson. Ga-easy does not imply steepest-ascent optimizable. In R. K. Belew and L. B. Booker, editors, Proc. of the 4th ICGA. mk, 1991.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Leila Kallel
    • 1
  1. 1.Ecole PolytechniqueCMAP - UMR CNRS 7641PalaiseauFrance

Personalised recommendations