Improving Identification of Difficult Small Classes by Balancing Class Distribution

@inproceedings{Laurikkala2001ImprovingIO,
  title={Improving Identification of Difficult Small Classes by Balancing Class Distribution},
  author={Jorma Laurikkala},
  booktitle={AIME},
  year={2001}
}
We studied three different methods to improve identification of small classes, which are also difficult to classify, by balancing imbalanced class distribution with data reduction. The new method, neighborhood cleaning rule (NCL), outperformed simple random selection within classes and one-sided selection method in experiments with ten real-world data sets. All reduction methods improved clearly identification of small classes (20-30%), but differences between the methods were insignificant… CONTINUE READING
Highly Influential
This paper has highly influenced 25 other papers. REVIEW HIGHLY INFLUENTIAL CITATIONS

From This Paper

Figures, tables, results, connections, and topics extracted from this paper.
185 Extracted Citations
18 Extracted References
Similar Papers

Citing Papers

Publications influenced by this paper.
Showing 1-10 of 185 extracted citations

Referenced Papers

Publications referenced by this paper.
Showing 1-10 of 18 references

UCI Repository of machine learning databases [http://www.ics.uci.edu/~mlearn/MLRepository.html

  • C. L. Blake, C. J. Merz
  • University of California, Department of…
  • 1998
2 Excerpts

Similar Papers

Loading similar papers…