Hypernetworks for Zero-shot Transfer in Reinforcement Learning.
Sahand Rezaei-ShoshtariCharlotte MorissetteFrancois Robert HoganGregory DudekDavid MegerPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- transfer learning
- knowledge transfer
- reinforcement learning algorithms
- function approximation
- state space
- control problems
- learning algorithm
- model free
- cross domain
- optimal control
- multi agent reinforcement learning
- markov decision processes
- learning problems
- temporal difference
- previously learned
- robotic control
- optimal policy
- object recognition
- learning classifier systems
- machine learning
- real world
- reinforcement learning methods
- data sets
- real time
- relational reinforcement learning
- database