Pairwise Naive Bayes Classifier
@inproceedings{Sulzmann2006PairwiseNB, title={Pairwise Naive Bayes Classifier}, author={Jan-Nikolas Sulzmann}, booktitle={LWA}, year={2006} }
Class binarizations are effective methods that break multi-class problem down into several 2class or binary problems to improve weak learners. This paper analyzes which effects these methods have if we choose a Naive Bayes learner for thebase classifier. Weconsider the knownunordered and pairwise class binarizations and propose an alternative approach for a pairwise calculation of a modified Naive Bayes classifier.
Topics from this paper
2 Citations
References
SHOWING 1-10 OF 13 REFERENCES
Probability Estimates for Multi-class Classification by Pairwise Coupling
- Mathematics, Computer Science
- J. Mach. Learn. Res.
- 2004
- 1,829
- PDF
Multi-Interval Discretization of Continuous-Valued Attributes for Classification Learning
- Computer Science
- IJCAI
- 1993
- 2,925
- PDF
D-64289
- Darmstadt, Germany, July
- 2006
Jan-Nikolas Sulzmann. Paarweiser Naive Bayes Klassifizierer
- Jan-Nikolas Sulzmann. Paarweiser Naive Bayes Klassifizierer
- 2006
Journal of Machine Learning Research (JMLR)
- 5:975–1005,
- 2004