A study of classification methods applying on KDD 2004 dataset

Che-Han Chang Chun-Wei Liu

National Taiwan University

Teaser

Abstract

We explore the KDD 2004 competition dataset and apply three classification methods base on supervised learning. We use different scales of training data, and different method to choose the proper hypothesis for each case. In the end, we have an in-class competition of performance on the dataset.

Download


Download the pdf file Paper:
4-page PDF

References

  1. “'Modest AdaBoost' - Teaching AdaBoost to Generalize Better.” [Link]
    Alexander Vezhnevets, and Vladimir Vezhnevets.
    2002
  2. “Applied Logistic Regression.” [Link]
    David W. Hosmer Jr., and Stanley Lemeshow.
    2000
  3. “LIBSVM: a library for support vector machines.” [Link]
    Chih-Chung Chang, and Chih-Jen Lin.
    2001