Generative Flow Networks for Precise Reward-Oriented Active Learning on Graphs.
Yinchuan LiZhigang LiWenqian LiYunfeng ShaoYan ZhengJianye HaoPublished in: IJCAI (2023)
Keyphrases
- active learning
- highly connected
- dynamic networks
- reinforcement learning
- average degree
- real world networks
- small world
- generative model
- fully connected
- social networks
- network analysis
- graph theory
- directed graph
- learning strategies
- training set
- relevance feedback
- graph theoretic
- semi supervised
- real world graphs
- labeled data
- graph structure
- graph matching
- community structure
- training examples
- complex networks
- network structure
- wireless networks
- data sets
- degree distribution
- selective sampling
- community discovery
- network size
- graph structures
- supervised learning
- text classification
- biological networks
- long run
- experimental design
- undirected graph
- spanning tree
- community detection
- semi supervised learning