A Transfer Approach Using Graph Neural Networks in Deep Reinforcement Learning.
Tianpei YangHeng YouJianye HaoYan ZhengMatthew E. TaylorPublished in: AAAI (2024)
Keyphrases
- neural network
- reinforcement learning
- transfer learning
- pattern recognition
- function approximation
- function approximators
- graph structure
- artificial neural networks
- graph theory
- graph representation
- genetic algorithm
- fuzzy logic
- connected components
- spanning tree
- graph theoretic
- directed graph
- back propagation
- random walk
- machine learning
- graph construction
- graph matching
- learning capabilities
- neural nets
- knn
- weighted graph
- state space
- knowledge transfer
- reinforcement learning algorithms
- evolutionary learning
- undirected graph
- directed acyclic graph
- multi layer
- optimal control
- multilayer perceptron
- markov decision processes
- optimal policy
- learning algorithm