Alleviating over-smoothing via graph sparsification based on vertex feature similarity.
Gongce WuShukuan LinYilin ZhuangJianzhong QiaoPublished in: Appl. Intell. (2023)
Keyphrases
- similarity graph
- directed graph
- undirected graph
- minimum weight
- hamiltonian cycle
- random walk
- labeled graphs
- similarity measure
- graph structure
- graph theory
- preprocessing step
- graph representation
- weighted graph
- strongly connected
- attributed graphs
- image features
- betweenness centrality
- vector representation
- feature vectors
- normalized cut
- edge weights
- graph model
- euclidean distance
- structured data
- distance measure
- preprocessing stage
- bipartite graph
- connected components
- vertex set
- distance function
- least squares
- average degree
- connected subgraphs
- inexact graph matching