Two Heads Are Better Than One: Boosting Graph Sparse Training via Semantic and Topological Awareness.
Guibin ZhangYanwei YueKun WangJunfeng FangYongduo SuiKai WangYuxuan LiangDawei ChengShirui PanTianlong ChenPublished in: CoRR (2024)
Keyphrases
- early stopping
- connected components
- graph model
- semantic information
- avoid overfitting
- semantic web
- gaussian graphical models
- graph representation
- sparse representation
- relevance vector machine
- training set
- directed graph
- topological map
- learning algorithm
- directed acyclic graph
- training process
- graph databases
- bipartite graph
- weighted graph
- ensemble methods
- semantic similarity
- high level
- machine learning
- neural network
- gaussian processes
- semantic network
- discriminative classifiers
- topological information
- supervised learning
- training data
- directed acyclic