Accelerating Reinforcement Learning by Composing Solutions of Automatically Identified Subtasks
Chris DrummondPublished in: CoRR (2011)
Keyphrases
- reinforcement learning
- neural network
- automatically generated
- state space
- markov decision processes
- database
- function approximation
- optimal solution
- benchmark problems
- state abstraction
- efficient solutions
- temporal difference
- learning tasks
- dynamic programming
- search algorithm
- learning environment
- decision making
- information retrieval
- machine learning
- real world