Proto-value functions: developmental reinforcement learning.
Sridhar MahadevanPublished in: ICML (2005)
Keyphrases
- reinforcement learning
- function approximation
- temporal difference
- reinforcement learning algorithms
- dynamic programming
- learning process
- machine learning
- data sets
- state space
- transfer learning
- multi agent reinforcement learning
- relational reinforcement learning
- action selection
- database
- robot control
- direct policy search
- transition model
- markov decision processes
- dynamic environments
- decision making
- information systems
- search engine
- artificial intelligence
- genetic algorithm
- data mining
- real time