The author introduces a novel approach, Hierarchical Multi-Relational Graph Representation Learning (HMGRL), to predict drug-drug interactions by capturing both explicit and implicit correlations between drugs. The approach leverages diverse data sources and spectral clustering to enhance prediction accuracy.
The author explores the potential of graph representation learning within a semi-supervised framework to predict fatty liver disease, emphasizing human-centric explanations through explainable GNNs.
薬物間相互作用の大規模予測のための階層的多関係グラフ表現学習
Proposing a novel method, GPCD, to enhance graph representation learning by extracting potential causes and mitigating label noise.
Introducing an architecture based on deep hierarchical decompositions to learn effective representations of large graphs.
Efficiently retrieve similar neural architectures using graph representation learning.
Enhancing the robustness and efficiency of Simplified PCNet for graph representation learning.
提案されたVQGRAPHフレームワークは、GNNからMLPへの知識蒸留において新しい最先端のパフォーマンスを達成します。
Large Language Models can effectively comprehend graph information through soft prompts, as demonstrated by the GraphPrompter framework.
STGED framework outperforms baselines in predicting future network connectivity for tactical communication networks.