Graph Structure Exploration for Reinforcement Learning State Embedding - Train Tetris Agent with Graph Neural Network.
Weijie GuanZhufeng LiHiroyuki YamauchiPublished in: ICECC (2023)
Keyphrases
- graph structure
- neural network
- reinforcement learning
- action selection
- graph structures
- exploration strategy
- graph model
- graphical models
- directed graph
- graph representations
- state action
- state space
- state abstraction
- weighted graph
- undirected graph
- agent receives
- multi agent
- action space
- autonomous learning
- directed acyclic
- tree structure
- agent learns
- learning agent
- graph representation
- graph theory
- information extraction
- learning algorithm