Applied Intelligence

, Volume 11, Issue 3, pp 277–284

Efficient GA Based Techniques for Classification

  • Peter K. Sharpe
  • Robin P. Glover

DOI: 10.1023/A:1008386925927

Cite this article as:
Sharpe, P.K. & Glover, R.P. Applied Intelligence (1999) 11: 277. doi:10.1023/A:1008386925927


A common approach to evaluating competing models in a classification context is via accuracy on a test set or on cross-validation sets. However, this can be computationally costly when using genetic algorithms with large datasets and the benefits of performing a wide search are compromised by the fact that estimates of the generalization abilities of competing models are subject to noise. This paper shows that clear advantages can be gained by using samples of the test set when evaluating competing models. Further, that applying statistical tests in combination with Occam's razor produces parsimonious models, matches the level of evaluation to the state of the search and retains the speed advantages of test set sampling.

genetic algorithmsclassificationdata mining

Copyright information

© Kluwer Academic Publishers 1999

Authors and Affiliations

  • Peter K. Sharpe
    • 1
  • Robin P. Glover
    • 1
  1. 1.Intelligent Computer Systems Centre, Faculty of Computer Studies and MathematicsUniversity of the West of EnglandBristol