A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection

Explore this paper's citation graph

Summary

The results indicate that for real-word datasets similar to the authors', the best method to use for model selection is ten fold stratified cross validation even if computation power allows using more folds.

Type
article
Published
1995-08-20
Cited by
13,926
References
26
Access
Open access

Keywords

Cross-validation, Naive Bayes classifier, Model selection, Computer science, Classifier (UML)

References

Cited by

Related papers