SimCalib: Graph Neural Network Calibration based on Similarity between Nodes.
Boshi TangZhiyong WuXixin WuQiaochu HuangJun ChenShun LeiHelen MengPublished in: CoRR (2023)
Keyphrases
- neural network
- directed graph
- graph structure
- adjacency matrix
- weighted graph
- undirected graph
- random walk
- small world
- average degree
- spanning tree
- similarity measure
- edge weights
- graph structures
- graph representation
- root node
- graph model
- pairwise similarities
- betweenness centrality
- dominating set
- nodes of a graph
- camera calibration
- artificial neural networks
- shortest path
- proximity measures
- random graphs
- finding the shortest path
- similarity graph
- strongly connected
- similarity scores
- path length
- network structure
- directed edges
- bipartite graph
- normalized cut
- pairwise
- graph matching
- graph theory
- structured data
- inexact graph matching
- connected graphs
- minimum cost
- planar graphs
- graph mining
- network model
- back propagation
- edit distance
- distance measure
- overlapping communities
- attributed graphs
- similarity function
- neural network model
- genetic algorithm
- citation graph
- distance function
- fully connected
- graph theoretic
- graph databases