Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171776673> ?p ?o ?g. }
- W3171776673 abstract "Exploiting symmetries and invariance in data is a powerful, yet not fully exploited, way to achieve better generalisation with more efficiency. In this paper, we introduce two graph network architectures that are equivariant to several types of transformations affecting the node coordinates. First, we build equivariance to any transformation in the coordinate embeddings that preserves the distance between neighbouring nodes, allowing for equivariance to the Euclidean group. Then, we introduce angle attributes to build equivariance to any angle preserving transformation - thus, to the conformal group. Thanks to their equivariance properties, the proposed models can be vastly more data efficient with respect to classical graph architectures, intrinsically equipped with a better inductive bias and better at generalising. We demonstrate these capabilities on a synthetic dataset composed of $n$-dimensional geometric objects. Additionally, we provide examples of their limitations when (the right) symmetries are not present in the data." @default.
- W3171776673 created "2021-06-22" @default.
- W3171776673 creator A5033706590 @default.
- W3171776673 creator A5081524121 @default.
- W3171776673 date "2021-05-28" @default.
- W3171776673 modified "2023-09-23" @default.
- W3171776673 title "Symmetry-driven graph neural networks." @default.
- W3171776673 cites W1493202709 @default.
- W3171776673 cites W1501856433 @default.
- W3171776673 cites W1508021176 @default.
- W3171776673 cites W1522301498 @default.
- W3171776673 cites W2116341502 @default.
- W3171776673 cites W2154579312 @default.
- W3171776673 cites W2279221249 @default.
- W3171776673 cites W2527189750 @default.
- W3171776673 cites W2541404351 @default.
- W3171776673 cites W2606780347 @default.
- W3171776673 cites W2788775653 @default.
- W3171776673 cites W2936357167 @default.
- W3171776673 cites W2949858440 @default.
- W3171776673 cites W2963464736 @default.
- W3171776673 cites W2996406751 @default.
- W3171776673 cites W2996443485 @default.
- W3171776673 cites W3007309629 @default.
- W3171776673 cites W3007332492 @default.
- W3171776673 cites W3007913393 @default.
- W3171776673 cites W3009578594 @default.
- W3171776673 cites W3034497530 @default.
- W3171776673 cites W3034650200 @default.
- W3171776673 cites W3034655405 @default.
- W3171776673 cites W3035547893 @default.
- W3171776673 cites W3035684957 @default.
- W3171776673 cites W3087257704 @default.
- W3171776673 cites W3102090196 @default.
- W3171776673 cites W3105320009 @default.
- W3171776673 cites W3118587469 @default.
- W3171776673 cites W3118691860 @default.
- W3171776673 cites W3118902573 @default.
- W3171776673 cites W3126960266 @default.
- W3171776673 cites W3131204112 @default.
- W3171776673 cites W3134275903 @default.
- W3171776673 cites W3144044690 @default.
- W3171776673 cites W3156762592 @default.
- W3171776673 cites W3157286395 @default.
- W3171776673 hasPublicationYear "2021" @default.
- W3171776673 type Work @default.
- W3171776673 sameAs 3171776673 @default.
- W3171776673 citedByCount "0" @default.
- W3171776673 crossrefType "posted-content" @default.
- W3171776673 hasAuthorship W3171776673A5033706590 @default.
- W3171776673 hasAuthorship W3171776673A5081524121 @default.
- W3171776673 hasConcept C104317684 @default.
- W3171776673 hasConcept C114614502 @default.
- W3171776673 hasConcept C120204988 @default.
- W3171776673 hasConcept C129782007 @default.
- W3171776673 hasConcept C132525143 @default.
- W3171776673 hasConcept C171036898 @default.
- W3171776673 hasConcept C173110770 @default.
- W3171776673 hasConcept C184720557 @default.
- W3171776673 hasConcept C185592680 @default.
- W3171776673 hasConcept C19939924 @default.
- W3171776673 hasConcept C202444582 @default.
- W3171776673 hasConcept C204241405 @default.
- W3171776673 hasConcept C2524010 @default.
- W3171776673 hasConcept C2993626508 @default.
- W3171776673 hasConcept C33923547 @default.
- W3171776673 hasConcept C41008148 @default.
- W3171776673 hasConcept C55493867 @default.
- W3171776673 hasConcept C80444323 @default.
- W3171776673 hasConcept C92757383 @default.
- W3171776673 hasConcept C96469262 @default.
- W3171776673 hasConcept C98214594 @default.
- W3171776673 hasConceptScore W3171776673C104317684 @default.
- W3171776673 hasConceptScore W3171776673C114614502 @default.
- W3171776673 hasConceptScore W3171776673C120204988 @default.
- W3171776673 hasConceptScore W3171776673C129782007 @default.
- W3171776673 hasConceptScore W3171776673C132525143 @default.
- W3171776673 hasConceptScore W3171776673C171036898 @default.
- W3171776673 hasConceptScore W3171776673C173110770 @default.
- W3171776673 hasConceptScore W3171776673C184720557 @default.
- W3171776673 hasConceptScore W3171776673C185592680 @default.
- W3171776673 hasConceptScore W3171776673C19939924 @default.
- W3171776673 hasConceptScore W3171776673C202444582 @default.
- W3171776673 hasConceptScore W3171776673C204241405 @default.
- W3171776673 hasConceptScore W3171776673C2524010 @default.
- W3171776673 hasConceptScore W3171776673C2993626508 @default.
- W3171776673 hasConceptScore W3171776673C33923547 @default.
- W3171776673 hasConceptScore W3171776673C41008148 @default.
- W3171776673 hasConceptScore W3171776673C55493867 @default.
- W3171776673 hasConceptScore W3171776673C80444323 @default.
- W3171776673 hasConceptScore W3171776673C92757383 @default.
- W3171776673 hasConceptScore W3171776673C96469262 @default.
- W3171776673 hasConceptScore W3171776673C98214594 @default.
- W3171776673 hasOpenAccess W3171776673 @default.
- W3171776673 hasRelatedWork W1575868470 @default.
- W3171776673 hasRelatedWork W2267339130 @default.
- W3171776673 hasRelatedWork W2590699358 @default.
- W3171776673 hasRelatedWork W2727408958 @default.
- W3171776673 hasRelatedWork W2745722324 @default.
- W3171776673 hasRelatedWork W2785253517 @default.