Feature Subset Selection Problems: A Variable-Length Chromosome Perspective

  • César M. Guerra-Salcedo


Fixed-length subset problems occur where a solution to a problem is described by an unordered subset of a particular cardinality. Some fixed-length subset problems are known as “deceptive” problems. In a deceptive problem certain possible solutions tend to lead the search algorithm towards a locally optimal solution. In this research, deceptive problems are treated as feature selection problems, the goal is to find the right combination of features that solves the problem. We explore a family of genetic search methods known as messy genetic algorithms on artificially generated deceptive problems and real-world instance-based classification problems.


Genetic Algorithm Selection Problem Probabilistic Neural Network Feature Selection Problem Traditional Genetic Algorithm 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Bankert, R. L., “Cloud Classification of avhrr Imagery in Maritime Regions Using a Probabilistic Neural Network”, Applied Metheorology, vol. 33(8), pp. 909–918, 1994.CrossRefGoogle Scholar
  2. [2]
    Crawford, K. D., “The Role of Recombination in Genetic Algorithms for Fixed-length Subset Problems”, PhD thesis, The Graduate School of The University of Tulsa, 1996.Google Scholar
  3. [3]
    Deb, K. and Goldberg, D., “Analyzing Deception in Trap Functions”, In Whitley, L. D., editor, FOGA-2, pp. 93–108, Morgan Kaufmann, 1993.Google Scholar
  4. [4]
    Eshelman, L., “The CHC Adaptive Search Algorithm. How to Have Safe Search When Engaging in Nontraditional Genetic Recombination”, In Rawlins, G., editor, FOGA-1, pp. 265–283, Morgan Kaufmann, 1991.Google Scholar
  5. [5]
    Goldberg, D. E., Deb, K., Kargupta, H., and Harik, G., “Rapid, Accurate Optimization of Difficult Problems Using Fast Messy Genetic Algorithms”, In Forrest, S., editor, Proc. of the 5th Int’l. Conf. on Gas, pp. 56–64, Morgan Kaufmann, 1993.Google Scholar
  6. [6]
    Guerra-Salcedo, C. and Whitley, D., “Genetic Search for Feature Subset Selection: A Comparison Between CHC and GENESIS”, In Proceedings of the Third Annual Genetic Programming Conference, pp. 504–509, Morgan Kaufmann, 1998.Google Scholar
  7. [7]
    Kargupta, H., “SEARCH, Polynomial Complexity, And The Fast Messy Genetic Algorithm”, PhD thesis, Department of Computer Science University of Illinois at Urbana Champaign, 1995.Google Scholar
  8. [8]
    Kohavi, R., “Wrappers for Performance Enhancement and Oblivious Decision Graphs”, PhD thesis, Stanford University, 1995.Google Scholar
  9. [9]
    Radcliffe, N. J. and George, F. A. W., “A Study In Set Recombination”, In Forrest, S., editor, Proc. of the 5th Int’l. Conf. on Gas, pp 23–30, Morgan Kaufmann, 1993.Google Scholar

Copyright information

© Springer-Verlag Wien 2001

Authors and Affiliations

  • César M. Guerra-Salcedo
    • 1
  1. 1.Advanced Technology GroupAthene Software Inc.BoulderUSA

Personalised recommendations