Unknown mixing times in apprenticeship and reinforcement learning.
Tom ZahavyAlon CohenHaim KaplanYishay MansourPublished in: UAI (2020)
Keyphrases
- reinforcement learning
- state space
- reinforcement learning algorithms
- function approximation
- initially unknown
- learning process
- markov decision processes
- neural network
- learning algorithm
- multi agent
- robotic control
- cognitive apprenticeship
- policy search
- optimal control
- orders of magnitude
- dynamic programming
- machine learning
- optimal policy
- evolutionary algorithm
- model free
- action selection
- expert systems
- partially observable
- robot control
- information systems
- information retrieval
- database