Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287123984> ?p ?o ?g. }
Showing items 1 to 63 of
63
with 100 items per page.
- W4287123984 abstract "In this paper, we investigate the question: Given a small number of datapoints, for example N = 30, how tight can PAC-Bayes and test set bounds be made? For such small datasets, test set bounds adversely affect generalisation performance by withholding data from the training procedure. In this setting, PAC-Bayes bounds are especially attractive, due to their ability to use all the data to simultaneously learn a posterior and bound its generalisation risk. We focus on the case of i.i.d. data with a bounded loss and consider the generic PAC-Bayes theorem of Germain et al. While their theorem is known to recover many existing PAC-Bayes bounds, it is unclear what the tightest bound derivable from their framework is. For a fixed learning algorithm and dataset, we show that the tightest possible bound coincides with a bound considered by Catoni; and, in the more natural case of distributions over datasets, we establish a lower bound on the best bound achievable in expectation. Interestingly, this lower bound recovers the Chernoff test set bound if the posterior is equal to the prior. Moreover, to illustrate how tight these bounds can be, we study synthetic one-dimensional classification tasks in which it is feasible to meta-learn both the prior and the form of the bound to numerically optimise for the tightest bounds possible. We find that in this simple, controlled scenario, PAC-Bayes bounds are competitive with comparable, commonly used Chernoff test set bounds. However, the sharpest test set bounds still lead to better guarantees on the generalisation error than the PAC-Bayes bounds we consider." @default.
- W4287123984 created "2022-07-25" @default.
- W4287123984 creator A5022158131 @default.
- W4287123984 creator A5023582352 @default.
- W4287123984 creator A5061327791 @default.
- W4287123984 creator A5072490106 @default.
- W4287123984 date "2021-06-07" @default.
- W4287123984 modified "2023-09-29" @default.
- W4287123984 title "How Tight Can PAC-Bayes be in the Small Data Regime?" @default.
- W4287123984 doi "https://doi.org/10.48550/arxiv.2106.03542" @default.
- W4287123984 hasPublicationYear "2021" @default.
- W4287123984 type Work @default.
- W4287123984 citedByCount "0" @default.
- W4287123984 crossrefType "posted-content" @default.
- W4287123984 hasAuthorship W4287123984A5022158131 @default.
- W4287123984 hasAuthorship W4287123984A5023582352 @default.
- W4287123984 hasAuthorship W4287123984A5061327791 @default.
- W4287123984 hasAuthorship W4287123984A5072490106 @default.
- W4287123984 hasBestOaLocation W42871239841 @default.
- W4287123984 hasConcept C107673813 @default.
- W4287123984 hasConcept C11413529 @default.
- W4287123984 hasConcept C114614502 @default.
- W4287123984 hasConcept C119857082 @default.
- W4287123984 hasConcept C134306372 @default.
- W4287123984 hasConcept C154945302 @default.
- W4287123984 hasConcept C169903167 @default.
- W4287123984 hasConcept C177264268 @default.
- W4287123984 hasConcept C199360897 @default.
- W4287123984 hasConcept C207201462 @default.
- W4287123984 hasConcept C33923547 @default.
- W4287123984 hasConcept C34388435 @default.
- W4287123984 hasConcept C41008148 @default.
- W4287123984 hasConcept C77553402 @default.
- W4287123984 hasConceptScore W4287123984C107673813 @default.
- W4287123984 hasConceptScore W4287123984C11413529 @default.
- W4287123984 hasConceptScore W4287123984C114614502 @default.
- W4287123984 hasConceptScore W4287123984C119857082 @default.
- W4287123984 hasConceptScore W4287123984C134306372 @default.
- W4287123984 hasConceptScore W4287123984C154945302 @default.
- W4287123984 hasConceptScore W4287123984C169903167 @default.
- W4287123984 hasConceptScore W4287123984C177264268 @default.
- W4287123984 hasConceptScore W4287123984C199360897 @default.
- W4287123984 hasConceptScore W4287123984C207201462 @default.
- W4287123984 hasConceptScore W4287123984C33923547 @default.
- W4287123984 hasConceptScore W4287123984C34388435 @default.
- W4287123984 hasConceptScore W4287123984C41008148 @default.
- W4287123984 hasConceptScore W4287123984C77553402 @default.
- W4287123984 hasLocation W42871239841 @default.
- W4287123984 hasOpenAccess W4287123984 @default.
- W4287123984 hasPrimaryLocation W42871239841 @default.
- W4287123984 hasRelatedWork W1996425479 @default.
- W4287123984 hasRelatedWork W2792951589 @default.
- W4287123984 hasRelatedWork W2977132600 @default.
- W4287123984 hasRelatedWork W3099765033 @default.
- W4287123984 hasRelatedWork W3185179407 @default.
- W4287123984 hasRelatedWork W3197848046 @default.
- W4287123984 hasRelatedWork W3201070945 @default.
- W4287123984 hasRelatedWork W4280583453 @default.
- W4287123984 hasRelatedWork W4286750367 @default.
- W4287123984 hasRelatedWork W4300029875 @default.
- W4287123984 isParatext "false" @default.
- W4287123984 isRetracted "false" @default.
- W4287123984 workType "article" @default.