VPE: Variational Policy Embedding for Transfer Reinforcement Learning.
Isac ArnekvistDanica KragicJohannes A. StorkPublished in: ICRA (2019)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- transfer learning
- action selection
- markov decision process
- state space
- function approximators
- partially observable environments
- action space
- reinforcement learning problems
- markov decision processes
- policy iteration
- state and action spaces
- actor critic
- partially observable
- reward function
- reinforcement learning algorithms
- markov decision problems
- control policy
- policy gradient
- control policies
- dynamic programming
- multi agent
- policy evaluation
- partially observable markov decision processes
- rl algorithms
- temporal difference
- approximate dynamic programming
- policy gradient methods
- model free
- function approximation
- image segmentation
- vector space
- decision problems
- average reward
- partially observable domains
- average cost
- inverse reinforcement learning
- learning algorithm
- knowledge transfer
- finite state
- long run
- infinite horizon
- neural network
- optical flow
- machine learning
- semi supervised
- continuous state spaces
- sufficient conditions
- learning tasks
- learning problems
- previously learned
- optimal control
- watermarking algorithm