Graph Constrained Reinforcement Learning for Natural Language Action Spaces.
Prithviraj AmmanabroluMatthew J. HausknechtPublished in: ICLR (2020)
Keyphrases
- action space
- reinforcement learning
- natural language
- state space
- state and action spaces
- markov decision processes
- continuous state
- real valued
- control policies
- action selection
- stochastic processes
- reinforcement learning methods
- function approximators
- continuous state spaces
- reinforcement learning problems
- machine learning
- skill learning
- function approximation
- reinforcement learning algorithms
- temporal difference
- random walk
- markov decision problems
- markov decision process
- dynamical systems
- shortest path
- markov chain
- dynamic programming
- search space
- multi agent
- learning algorithm