Grouping of Maintenance Actions with Deep Reinforcement Learning and Graph Convolutional Networks.
David KerkkampZaharah Allah BukhshYingqian ZhangNils JansenPublished in: ICAART (2) (2022)
Keyphrases
- reinforcement learning
- deep learning
- action selection
- perceptual aliasing
- graph model
- average degree
- partially observable domains
- complex networks
- graph theory
- state and action spaces
- edge weights
- graph partitioning
- action space
- random walk
- bipartite graph
- graph structure
- graph mining
- directed graph
- social networks
- markov decision processes
- partially observable
- fully connected
- reward function
- supervised learning
- reinforcement learning algorithms
- learning algorithm
- graph structures
- state action
- community discovery
- multiagent reinforcement learning
- betweenness centrality
- small world
- graph representation
- function approximation
- structured data
- biological networks
- model free
- partial observability
- citation networks
- decision theoretic
- network structure
- densely connected
- state space
- graph layout
- machine learning
- unsupervised feature learning