On Designing Pattern Classifiers Using Artificially Created Bootstrap Samples

Qun Wang, B. John Oommen

2004

Abstract

We consider the problem of building a pattern recognition classifier using a set of training samples. Traditionally, the classifier is constructed by using only the set1 of given training samples. But the quality of the classifier is poor when the size of the training sample is small. In this paper, we shall show that the quality of the classifier can be improved by utilizing artificially created training samples, where the latter are obtained by using various extensions of Efron’s bootstrap technique. Experimental results show that classifiers which incorporate some of the bootstrap algorithms, noticeably improve the performance of the resultant classifier.

References

  1. Chernick, M.R., Murthy, V.K. and Nealy, C.D. (1985), “Application of bootstrap and other resampling techniques: evaluation of classifier performance”, Pattern Recognition Letters, 3, pp.167 - 178.
  2. Davison, A. C. and Hall, P. (1992), “On the bias and variability of bootstrap and cross-validation estimates of error rate in discrimination problems”, Biometrika, 79, pp.279 - 284.
  3. Duda, R.O, Hart, P. E. and Stork, D. G., (2001), Pattern Classification, Second Edition, Wiley & Sons (2001).
  4. Efron, B. (1979), “Bootstrap Method: Another Look at the Jackknife”, Annals of Statistics, 7, pp.1 - 26.
  5. Efron, B. (1983), “Estimating the error rate of a prediction rule: improvment on cross-validation”, Journal of the American Statistical Association, 78, pp.316 - 331.
  6. Efron, B. and Tibshirani, R. J. (1997), “Improvement on cross-validation: the .632 + bootstap method”, Journal of the American Statistical Association, 92, pp.548 - 560.
  7. Hand, J.D., (1986), “Recent advances in error rate estimation”, Pattern Recognition Letters, 4, pp.335 - 346.
  8. Jain, A.K., Dubes, R.C. and Chen, C. (1987), “Bootstrap techniques for error estimation”, IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-9, 628 - 633.
  9. Taylor, M. S. and Thompson, J. R., (1992), “A nonparametric density estimation based resampling algorithm”, Exploring the Limits of Bootstrap, John Wiley & Sons, Inc., pp.397 - 404.
  10. Wang, Q. and Oommen, B. J. (2003), “Classification Error-Rate Estimation Using New Pseudo-Sample Bootstrap Methods”, Pattern Recognition in Information System, PRIS-2003, Jean-Marc Ogier and Eric Trupin (Eds.), pp.
  11. Wang, Q. (2000), “Bootstrap Techniques for Statistical Pattern Recognition”, Master Thesis, Carleton University.
  12. Zhen, Z. (1987), “Random weighting methods”, Acta Math. Appl. Sinica, 10, pp.247 - 253.
Download


Paper Citation


in Harvard Style

Wang Q. and John Oommen B. (2004). On Designing Pattern Classifiers Using Artificially Created Bootstrap Samples . In Proceedings of the 4th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2004) ISBN 972-8865-01-5, pages 159-168. DOI: 10.5220/0002653901590168


in Bibtex Style

@conference{pris04,
author={Qun Wang and B. John Oommen},
title={On Designing Pattern Classifiers Using Artificially Created Bootstrap Samples},
booktitle={Proceedings of the 4th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2004)},
year={2004},
pages={159-168},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002653901590168},
isbn={972-8865-01-5},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 4th International Workshop on Pattern Recognition in Information Systems - Volume 1: PRIS, (ICEIS 2004)
TI - On Designing Pattern Classifiers Using Artificially Created Bootstrap Samples
SN - 972-8865-01-5
AU - Wang Q.
AU - John Oommen B.
PY - 2004
SP - 159
EP - 168
DO - 10.5220/0002653901590168