Matches in SemOpenAlex for { <https://semopenalex.org/work/W4380993417> ?p ?o ?g. }
Showing items 1 to 53 of
53
with 100 items per page.
- W4380993417 abstract "Self-supervised learning has shown its promising capability in graph representation learning in recent work. Most existing pre-training strategies usually choose the popular Graph neural networks (GNNs), which can be seen as a special form of low-pass filter, fail to effectively capture heterophily. In this paper, we first present an experimental investigation exploring the performance of low-pass and high-pass filters in heterophily graph classification, where the results clearly show that high-frequency signal is important for learning heterophily graph representation. On the other hand, it is still unclear how to effectively capture the structural pattern of graphs and how to measure the capability of the self-supervised pre-training strategy in capturing graph structure. To address the problem, we first design a quantitative metric to Measure Graph Structure (MGS), which analyzes correlation between structural similarity and embedding similarity of graph pairs. Then, to enhance the graph structural information captured by self-supervised learning, we propose a novel self-supervised strategy for Pre-training GNNs based on the Metric (PGM). Extensive experiments validate our pre-training strategy achieves state-of-the-art performance for molecular property prediction and protein function prediction. In addition, we find choosing the suitable filter sometimes may be better than designing good pre-training strategies for heterophily graph classification." @default.
- W4380993417 created "2023-06-17" @default.
- W4380993417 creator A5055040307 @default.
- W4380993417 creator A5058479146 @default.
- W4380993417 creator A5060839385 @default.
- W4380993417 date "2023-06-14" @default.
- W4380993417 modified "2023-09-26" @default.
- W4380993417 title "Self-supervised Learning and Graph Classification under Heterophily" @default.
- W4380993417 doi "https://doi.org/10.48550/arxiv.2306.08469" @default.
- W4380993417 hasPublicationYear "2023" @default.
- W4380993417 type Work @default.
- W4380993417 citedByCount "0" @default.
- W4380993417 crossrefType "posted-content" @default.
- W4380993417 hasAuthorship W4380993417A5055040307 @default.
- W4380993417 hasAuthorship W4380993417A5058479146 @default.
- W4380993417 hasAuthorship W4380993417A5060839385 @default.
- W4380993417 hasBestOaLocation W43809934171 @default.
- W4380993417 hasConcept C119857082 @default.
- W4380993417 hasConcept C132525143 @default.
- W4380993417 hasConcept C153180895 @default.
- W4380993417 hasConcept C154945302 @default.
- W4380993417 hasConcept C2776517306 @default.
- W4380993417 hasConcept C41008148 @default.
- W4380993417 hasConcept C41608201 @default.
- W4380993417 hasConcept C58973888 @default.
- W4380993417 hasConcept C75564084 @default.
- W4380993417 hasConcept C80444323 @default.
- W4380993417 hasConceptScore W4380993417C119857082 @default.
- W4380993417 hasConceptScore W4380993417C132525143 @default.
- W4380993417 hasConceptScore W4380993417C153180895 @default.
- W4380993417 hasConceptScore W4380993417C154945302 @default.
- W4380993417 hasConceptScore W4380993417C2776517306 @default.
- W4380993417 hasConceptScore W4380993417C41008148 @default.
- W4380993417 hasConceptScore W4380993417C41608201 @default.
- W4380993417 hasConceptScore W4380993417C58973888 @default.
- W4380993417 hasConceptScore W4380993417C75564084 @default.
- W4380993417 hasConceptScore W4380993417C80444323 @default.
- W4380993417 hasLocation W43809934171 @default.
- W4380993417 hasOpenAccess W4380993417 @default.
- W4380993417 hasPrimaryLocation W43809934171 @default.
- W4380993417 hasRelatedWork W2893186803 @default.
- W4380993417 hasRelatedWork W2923818335 @default.
- W4380993417 hasRelatedWork W3035116611 @default.
- W4380993417 hasRelatedWork W3044354590 @default.
- W4380993417 hasRelatedWork W3149439221 @default.
- W4380993417 hasRelatedWork W4212923699 @default.
- W4380993417 hasRelatedWork W4213133066 @default.
- W4380993417 hasRelatedWork W4285120483 @default.
- W4380993417 hasRelatedWork W4287763734 @default.
- W4380993417 hasRelatedWork W4323323198 @default.
- W4380993417 isParatext "false" @default.
- W4380993417 isRetracted "false" @default.
- W4380993417 workType "article" @default.