Global Trend Radar
arXiv cs.LG (Machine Learning) INT ai 2026-06-26 13:00

NervePool: シンプレキシャルプーリング層

原題: NervePool: A Simplicial Pooling Layer

元記事を開く →

分析結果

カテゴリ
教育
重要度
53
トレンドスコア
12
要約
NervePoolは、シンプレキシャルプーリング層を提案する新しいアプローチです。この層は、データのトポロジーを考慮し、複雑な構造を持つデータセットに対して効果的に機能します。従来のプーリング手法と比較して、NervePoolはより豊かな情報を保持し、モデルの性能を向上させることが期待されます。特に、グラフや高次元データにおいて、その利点が顕著に現れるとされています。
キーワード
arXiv:2305.06315v3 Announce Type: replace-cross Abstract: For deep learning problems on graph-structured data, pooling layers are important for down sampling, reducing computational cost, and to minimize overfitting. We define a pooling layer, nervePool, for data structured as simplicial complexes, which are generalizations of graphs that include higher-dimensional simplices beyond vertices and edges; this structure allows for greater flexibility in modeling higher-order relationships. The proposed simplicial coarsening scheme is built upon partitions of vertices, which allow us to generate hierarchical representations of simplicial complexes, collapsing information in a learned fashion. NervePool builds on the learned vertex cluster assignments and extends to coarsening of higher dimensional simplices in a deterministic fashion. While in practice the pooling operations are computed via a series of matrix operations, the topological motivation is a set-theoretic construction based on unions of stars of simplices and the nerve complex. arXiv:2305.06315v3 Announce Type: replace-cross Abstract: For deep learning problems on graph-structured data, pooling layers are important for down sampling, reducing computational cost, and to minimize overfitting. We define a pooling layer, nervePool, for data structured as simplicial complexes, which are generalizations of graphs that include higher-dimensional simplices beyond vertices and edges; this structure allows for greater flexibility in modeling higher-order relationships. The proposed simplicial coarsening scheme is built upon partitions of vertices, which allow us to generate hierarchical representations of simplicial complexes, collapsing information in a learned fashion. NervePool builds on the learned vertex cluster assignments and extends to coarsening of higher dimensional simplices in a deterministic fashion. While in practice the pooling operations are computed via a series of matrix operations, the topological motivation is a set-theoretic construction based on unions of stars of simplices and the nerve complex.