| Reference : Aggregated hold-out |
| Scientific journals : Article | |||
| Physical, chemical, mathematical & earth Sciences : Mathematics | |||
| http://hdl.handle.net/10993/47480 | |||
| Aggregated hold-out | |
| English | |
Maillard, Guillaume  [University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Mathematics (DMATH) >] | |
| Arlot, Sylvain [Université Paris-Saclay > Mathematics > > Professor] | |
| Lerasle, Matthieu [ENSAE > > > CR] | |
| Jan-2021 | |
| Journal of Machine Learning Research | |
| MIT Press | |
| 22 | |
| Yes (verified by ORBilu) | |
| International | |
| 1532-4435 | |
| 1533-7928 | |
| Brookline | |
| MA | |
| [en] cross-validation ; aggregation ; bagging ; hyperparameter selection ; regularized kernel regression | |
| [en] Aggregated hold-out (agghoo) is a method which averages learning rules selected by hold-out
(that is, cross-validation with a single split). We provide the first theoretical guarantees on agghoo, ensuring that it can be used safely: Agghoo performs at worst like the hold-out when the risk is convex. The same holds true in classification with the 0--1 risk, with an additional constant factor. For the hold-out, oracle inequalities are known for bounded losses, as in binary classification. We show that similar results can be proved, under appropriate assumptions, for other risk-minimization problems. In particular, we obtain an oracle inequality for regularized kernel regression with a Lipschitz loss, without requiring that the $Y$ variable or the regressors be bounded. Numerical experiments show that aggregation brings a significant improvement over the hold-out and that agghoo is competitive with cross-validation. | |
| Université Paris-Sud, University of Luxembourg | |
| European Union Horizon 2020 | |
| Researchers | |
| http://hdl.handle.net/10993/47480 |
| File(s) associated to this reference | ||||||||||||||
|
Fulltext file(s):
| ||||||||||||||
All documents in ORBilu are protected by a user license.
