Abstract
In order to calibrate a penalization procedure for model selection, the statistician has to choose a shape for the penalty and a leading constant. In this paper, we study, for the marginal density estimation problem, the resampling penalties as general estimators of the shape of an ideal penalty. We prove that the selected estimator satisfies sharp oracle inequalities without remainder terms under a few assumptions on the marginal density s and the collection of models. We also study the slope heuristic, which yields a data-driven choice of the leading constant in front of the penalty when the complexity of the models is well-chosen.
| Original language | English |
|---|---|
| Pages (from-to) | 884-908 |
| Number of pages | 25 |
| Journal | Annales de l'institut Henri Poincare (B) Probability and Statistics |
| Volume | 48 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 1 Aug 2012 |
| Externally published | Yes |
Keywords
- Density estimation
- Optimal model selection
- Resampling methods
- Slope heuristic
Fingerprint
Dive into the research topics of 'Optimal model selection in density estimation'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver