Matches in SemOpenAlex for { <https://semopenalex.org/work/W3172205380> ?p ?o ?g. }
- W3172205380 endingPage "11227" @default.
- W3172205380 startingPage "11217" @default.
- W3172205380 abstract "Recent observations have advanced our understanding of the neural network optimization landscape, revealing the existence of (1) paths of high accuracy containing diverse solutions and (2) wider minima offering improved performance. Previous methods observing diverse paths require multiple training runs. In contrast we aim to leverage both property (1) and (2) with a single method and in a single training run. With a similar computational cost as training one model, we learn lines, curves, and simplexes of high-accuracy neural networks. These neural network subspaces contain diverse solutions that can be ensembled, approaching the ensemble performance of independently trained networks without the training cost. Moreover, using the subspace midpoint boosts accuracy, calibration, and robustness to label noise, outperforming Stochastic Weight Averaging." @default.
- W3172205380 created "2021-06-22" @default.
- W3172205380 creator A5001590363 @default.
- W3172205380 creator A5012428670 @default.
- W3172205380 creator A5026171268 @default.
- W3172205380 creator A5056246621 @default.
- W3172205380 creator A5090739892 @default.
- W3172205380 date "2021-07-18" @default.
- W3172205380 modified "2023-09-26" @default.
- W3172205380 title "Learning Neural Network Subspaces" @default.
- W3172205380 hasPublicationYear "2021" @default.
- W3172205380 type Work @default.
- W3172205380 sameAs 3172205380 @default.
- W3172205380 citedByCount "3" @default.
- W3172205380 countsByYear W31722053802021 @default.
- W3172205380 crossrefType "proceedings-article" @default.
- W3172205380 hasAuthorship W3172205380A5001590363 @default.
- W3172205380 hasAuthorship W3172205380A5012428670 @default.
- W3172205380 hasAuthorship W3172205380A5026171268 @default.
- W3172205380 hasAuthorship W3172205380A5056246621 @default.
- W3172205380 hasAuthorship W3172205380A5090739892 @default.
- W3172205380 hasConcept C104317684 @default.
- W3172205380 hasConcept C105795698 @default.
- W3172205380 hasConcept C111472728 @default.
- W3172205380 hasConcept C11413529 @default.
- W3172205380 hasConcept C119857082 @default.
- W3172205380 hasConcept C12362212 @default.
- W3172205380 hasConcept C126255220 @default.
- W3172205380 hasConcept C134306372 @default.
- W3172205380 hasConcept C138885662 @default.
- W3172205380 hasConcept C144521790 @default.
- W3172205380 hasConcept C153083717 @default.
- W3172205380 hasConcept C153180895 @default.
- W3172205380 hasConcept C154945302 @default.
- W3172205380 hasConcept C185429906 @default.
- W3172205380 hasConcept C185592680 @default.
- W3172205380 hasConcept C186633575 @default.
- W3172205380 hasConcept C189950617 @default.
- W3172205380 hasConcept C2524010 @default.
- W3172205380 hasConcept C32834561 @default.
- W3172205380 hasConcept C33923547 @default.
- W3172205380 hasConcept C34130140 @default.
- W3172205380 hasConcept C41008148 @default.
- W3172205380 hasConcept C41045048 @default.
- W3172205380 hasConcept C50644808 @default.
- W3172205380 hasConcept C55493867 @default.
- W3172205380 hasConcept C62438384 @default.
- W3172205380 hasConcept C63479239 @default.
- W3172205380 hasConceptScore W3172205380C104317684 @default.
- W3172205380 hasConceptScore W3172205380C105795698 @default.
- W3172205380 hasConceptScore W3172205380C111472728 @default.
- W3172205380 hasConceptScore W3172205380C11413529 @default.
- W3172205380 hasConceptScore W3172205380C119857082 @default.
- W3172205380 hasConceptScore W3172205380C12362212 @default.
- W3172205380 hasConceptScore W3172205380C126255220 @default.
- W3172205380 hasConceptScore W3172205380C134306372 @default.
- W3172205380 hasConceptScore W3172205380C138885662 @default.
- W3172205380 hasConceptScore W3172205380C144521790 @default.
- W3172205380 hasConceptScore W3172205380C153083717 @default.
- W3172205380 hasConceptScore W3172205380C153180895 @default.
- W3172205380 hasConceptScore W3172205380C154945302 @default.
- W3172205380 hasConceptScore W3172205380C185429906 @default.
- W3172205380 hasConceptScore W3172205380C185592680 @default.
- W3172205380 hasConceptScore W3172205380C186633575 @default.
- W3172205380 hasConceptScore W3172205380C189950617 @default.
- W3172205380 hasConceptScore W3172205380C2524010 @default.
- W3172205380 hasConceptScore W3172205380C32834561 @default.
- W3172205380 hasConceptScore W3172205380C33923547 @default.
- W3172205380 hasConceptScore W3172205380C34130140 @default.
- W3172205380 hasConceptScore W3172205380C41008148 @default.
- W3172205380 hasConceptScore W3172205380C41045048 @default.
- W3172205380 hasConceptScore W3172205380C50644808 @default.
- W3172205380 hasConceptScore W3172205380C55493867 @default.
- W3172205380 hasConceptScore W3172205380C62438384 @default.
- W3172205380 hasConceptScore W3172205380C63479239 @default.
- W3172205380 hasLocation W31722053801 @default.
- W3172205380 hasOpenAccess W3172205380 @default.
- W3172205380 hasPrimaryLocation W31722053801 @default.
- W3172205380 hasRelatedWork W1481935768 @default.
- W3172205380 hasRelatedWork W1915968771 @default.
- W3172205380 hasRelatedWork W2609644570 @default.
- W3172205380 hasRelatedWork W2612983688 @default.
- W3172205380 hasRelatedWork W2739997491 @default.
- W3172205380 hasRelatedWork W2890311316 @default.
- W3172205380 hasRelatedWork W2896230333 @default.
- W3172205380 hasRelatedWork W2905801789 @default.
- W3172205380 hasRelatedWork W2914317573 @default.
- W3172205380 hasRelatedWork W2925974453 @default.
- W3172205380 hasRelatedWork W2946742785 @default.
- W3172205380 hasRelatedWork W2971119958 @default.
- W3172205380 hasRelatedWork W3006614220 @default.
- W3172205380 hasRelatedWork W3009185440 @default.
- W3172205380 hasRelatedWork W3034538858 @default.
- W3172205380 hasRelatedWork W3044287261 @default.
- W3172205380 hasRelatedWork W3094364451 @default.
- W3172205380 hasRelatedWork W3129819231 @default.
- W3172205380 hasRelatedWork W3130422950 @default.
- W3172205380 hasRelatedWork W3204818005 @default.