Genetic Algorithm Selection of Features for Hand-printed Character Identification

  • Roger S. Gaborski
  • Peter G. Anderson
  • Christopher T. Asbury
  • David G. Tilley


We have constructed a linear discriminator for handprinted character recognition that uses a (binary) vector of 1, 500 features based on an equidistributed collection of products of pixel pairs. This classifier is competitive with other techniques, but faster to train and to run for classification.

However, the 1, 500-member feature set clearly contains many redundant (overlapping or useless) members, and a significantly smaller set would be very desirable (e.g., for faster training, a faster and smaller application program, and a smaller system suitable for hardware implementation). A system using the small set of features should also be better at generalization, since fewer features are less likely to allow a system to “memorize noise in the training data.”

We tried several genetic algorithm approaches to search for effective small subsets of features, and we have successfully found a 300-element set of features and built a classifier whose performance is as good on our testing set as the system using the full feature set.


Genetic Algorithm Feature Subset Pixel Pair Genetic Search Training Exemplar 
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]
    Peter G. Anderson and Roger S. Gaborski, “The polynomial method augmented by supervised training for hand printed character recognition,” Proceedings of the International Conference on Neural Networks and Genetic Algorithms, 1993.Google Scholar
  2. [2]
    Frank Z. Brill, Donald E. Brown, and Worthy N. Martin, “Fast genetic selection of features for neural network classifiers,” IEEE Transactions on Neural Networks, Vol. 3, No. 2, March 1992.Google Scholar
  3. [3]
    Eric I. Chang, Richard P. Lippmann, and David W. Tong, “Using genetic algorithms to select and create features for pattern classification,” Proceedings of the International Joint Conference on Neural Networks, 1990.Google Scholar
  4. [4]
    Lawrence Davis (ed.), Handbook of Genetic Algorithms, Van Nostrand Reinhold, New York, 1991.Google Scholar
  5. [5]
    D. E. Goldberg, Genetic Algorithms in Search, Optimization, and Machine Learning, Addison-Wesley, New York, 1989.MATHGoogle Scholar
  6. [6]
    A. Rao, P. G. Anderson, R. S. Gaborski, and K. S. Jaiswal, “A hardware polynomial feature net for handprinted digit recognition,” Proceedings of the Third IEE International Conference on Artificial Neural Networks, 1993.Google Scholar
  7. [7]
    Uma Shrinivasan, “Polynomial discriminant method for handwritten digit recognition,” SUNY Buffalo Technical Report, December 14, 1989.Google Scholar
  8. [8]
    W. Siedlecki and J. Sklansky, “Constrained genetic optimization via reward-penalty balancing and its use in pattern recognition,” Proceedings of the Third International Conference on Genetic Algorithms, 1989.Google Scholar
  9. [9]
    Darreil Whitley, “The GENITOR algorithm and selective pressure: why rank-based allocation of reproductive trials is best,” Proceedings of the International Conference on Genetic Algorithms, 1989.Google Scholar

Copyright information

© Springer-Verlag/Wien 1993

Authors and Affiliations

  • Roger S. Gaborski
    • 1
  • Peter G. Anderson
    • 1
  • Christopher T. Asbury
    • 1
  • David G. Tilley
    • 1
  1. 1.Imaging Research LaboratoriesEastman Kodak CompanyRochesterUSA

Personalised recommendations