Searching for Plannable Domains can Speed up Reinforcement Learning
Istvan SzitaBálint TakácsAndrás LörinczPublished in: CoRR (2002)
Keyphrases
- reinforcement learning
- transfer learning
- function approximation
- real world
- markov decision processes
- application domains
- partially observable domains
- real time
- deterministic domains
- complex domains
- machine learning
- search strategies
- learning algorithm
- model free
- cross domain
- state space
- databases
- data sets
- neural network
- semi supervised
- temporal difference
- reinforcement learning algorithms
- multi agent
- robotic control
- transferring knowledge
- case study