Policy Generation from Latent Embeddings for Reinforcement Learning.
Corentin ArtaudRafael PinaXiyu ShiVaruna De-SilvaPublished in: ISPR (2) (2023)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- action selection
- function approximation
- actor critic
- reinforcement learning algorithms
- policy evaluation
- state space
- policy iteration
- partially observable domains
- policy gradient
- partially observable environments
- control policies
- action space
- machine learning
- low dimensional
- continuous state spaces
- latent variables
- markov decision processes
- model free
- approximate dynamic programming
- reward function
- partially observable
- manifold learning
- reinforcement learning problems
- control problems
- state and action spaces
- markov decision problems
- temporal difference
- euclidean space
- multi agent
- vector space
- function approximators
- rl algorithms
- state dependent
- latent space
- long run
- optimal control
- decision problems
- inverse reinforcement learning
- exploration exploitation tradeoff
- policy gradient methods
- dimensionality reduction
- transfer learning
- transition model
- generation process
- control policy