Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376876721> ?p ?o ?g. }
Showing items 1 to 61 of
61
with 100 items per page.
- W4376876721 abstract "Node classification tasks on graphs are addressed via fully-trained deep message-passing models that learn a hierarchy of node representations via multiple aggregations of a node's neighbourhood. While effective on graphs that exhibit a high ratio of intra-class edges, this approach poses challenges in the opposite case, i.e. heterophily, where nodes belonging to the same class are usually further apart. In graphs with a high degree of heterophily, the smoothed representations based on close neighbours computed by convolutional models are no longer effective. So far, architectural variations in message-passing models to reduce excessive smoothing or rewiring the input graph to improve longer-range message passing have been proposed. In this paper, we address the challenges of heterophilic graphs with Graph Echo State Network (GESN) for node classification. GESN is a reservoir computing model for graphs, where node embeddings are recursively computed by an untrained message-passing function. Our experiments show that reservoir models are able to achieve better or comparable accuracy with respect to most fully trained deep models that implement ad hoc variations in the architectural bias or perform rewiring as a preprocessing step on the input graph, with an improvement in terms of efficiency/accuracy trade-off. Furthermore, our analysis shows that GESN is able to effectively encode the structural relationships of a graph node, by showing a correlation between iterations of the recursive embedding function and the distribution of shortest paths in a graph." @default.
- W4376876721 created "2023-05-18" @default.
- W4376876721 creator A5010923663 @default.
- W4376876721 creator A5058183468 @default.
- W4376876721 date "2023-05-14" @default.
- W4376876721 modified "2023-09-29" @default.
- W4376876721 title "Addressing Heterophily in Node Classification with Graph Echo State Networks" @default.
- W4376876721 doi "https://doi.org/10.48550/arxiv.2305.08233" @default.
- W4376876721 hasPublicationYear "2023" @default.
- W4376876721 type Work @default.
- W4376876721 citedByCount "0" @default.
- W4376876721 crossrefType "posted-content" @default.
- W4376876721 hasAuthorship W4376876721A5010923663 @default.
- W4376876721 hasAuthorship W4376876721A5058183468 @default.
- W4376876721 hasBestOaLocation W43768767211 @default.
- W4376876721 hasConcept C104317684 @default.
- W4376876721 hasConcept C11413529 @default.
- W4376876721 hasConcept C120314980 @default.
- W4376876721 hasConcept C127413603 @default.
- W4376876721 hasConcept C132525143 @default.
- W4376876721 hasConcept C154945302 @default.
- W4376876721 hasConcept C185592680 @default.
- W4376876721 hasConcept C34736171 @default.
- W4376876721 hasConcept C41008148 @default.
- W4376876721 hasConcept C55493867 @default.
- W4376876721 hasConcept C62611344 @default.
- W4376876721 hasConcept C66746571 @default.
- W4376876721 hasConcept C66938386 @default.
- W4376876721 hasConcept C80444323 @default.
- W4376876721 hasConcept C854659 @default.
- W4376876721 hasConceptScore W4376876721C104317684 @default.
- W4376876721 hasConceptScore W4376876721C11413529 @default.
- W4376876721 hasConceptScore W4376876721C120314980 @default.
- W4376876721 hasConceptScore W4376876721C127413603 @default.
- W4376876721 hasConceptScore W4376876721C132525143 @default.
- W4376876721 hasConceptScore W4376876721C154945302 @default.
- W4376876721 hasConceptScore W4376876721C185592680 @default.
- W4376876721 hasConceptScore W4376876721C34736171 @default.
- W4376876721 hasConceptScore W4376876721C41008148 @default.
- W4376876721 hasConceptScore W4376876721C55493867 @default.
- W4376876721 hasConceptScore W4376876721C62611344 @default.
- W4376876721 hasConceptScore W4376876721C66746571 @default.
- W4376876721 hasConceptScore W4376876721C66938386 @default.
- W4376876721 hasConceptScore W4376876721C80444323 @default.
- W4376876721 hasConceptScore W4376876721C854659 @default.
- W4376876721 hasLocation W43768767211 @default.
- W4376876721 hasOpenAccess W4376876721 @default.
- W4376876721 hasPrimaryLocation W43768767211 @default.
- W4376876721 hasRelatedWork W1568080662 @default.
- W4376876721 hasRelatedWork W2356694558 @default.
- W4376876721 hasRelatedWork W2389073067 @default.
- W4376876721 hasRelatedWork W2971815656 @default.
- W4376876721 hasRelatedWork W2976903742 @default.
- W4376876721 hasRelatedWork W3033495140 @default.
- W4376876721 hasRelatedWork W4288102844 @default.
- W4376876721 hasRelatedWork W4288238488 @default.
- W4376876721 hasRelatedWork W4317655900 @default.
- W4376876721 hasRelatedWork W4372260176 @default.
- W4376876721 isParatext "false" @default.
- W4376876721 isRetracted "false" @default.
- W4376876721 workType "article" @default.