ML Anthology

Suboptimal Behavior of Bayes and MDL in Classification Under Misspecification

Peter Grünwald, John Langford
COLT 2004 pp. 331-347
doi:10.1007/978-3-540-27819-1_23 /colt/2004/grunwald2004colt-suboptimal/

Abstract

We show that forms of Bayesian and MDL inference that are often applied to classification problems can be inconsistent . This means there exists a learning problem such that for all amounts of data the generalization errors of the MDL classifier and the Bayes classifier relative to the Bayesian posterior both remain bounded away from the smallest achievable generalization error.

PDF COLT Semantic Scholar

Cite

Text

Grünwald and Langford. "Suboptimal Behavior of Bayes and MDL in Classification Under Misspecification." Annual Conference on Computational Learning Theory, 2004. doi:10.1007/978-3-540-27819-1_23

Markdown

[Grünwald and Langford. "Suboptimal Behavior of Bayes and MDL in Classification Under Misspecification." Annual Conference on Computational Learning Theory, 2004.](https://mlanthology.org/colt/2004/grunwald2004colt-suboptimal/) doi:10.1007/978-3-540-27819-1_23

BibTeX

@inproceedings{grunwald2004colt-suboptimal,
  title     = {{Suboptimal Behavior of Bayes and MDL in Classification Under Misspecification}},
  author    = {Grünwald, Peter and Langford, John},
  booktitle = {Annual Conference on Computational Learning Theory},
  year      = {2004},
  pages     = {331-347},
  doi       = {10.1007/978-3-540-27819-1_23},
  url       = {https://mlanthology.org/colt/2004/grunwald2004colt-suboptimal/}
}