articleJournal of Machine Learning ResearchDec 1, 2004Closed access

In Defense of One-Vs-All Classification

RRifkinRyanKKlautauAldebaro

Abstract

We consider the problem of multiclass classification. Our main thesis is that a simple one-vs-all scheme is as accurate as any other approach, assuming that the underlying binary classifiers are ...

Citation impact

897
total citations
FWCI
29.14
Percentile
100%
References
0
Citations per year

Authors

2

Topics & keywords

Topics
Keywords
  • Multiclass classification
  • Simple (philosophy)
  • Classification scheme
  • Scheme (mathematics)
  • Artificial intelligence
  • Binary classification
  • Computer science
  • Binary number
UN Sustainable Development Goals
  • Reduced inequalities
No related works found for this paper.