Memory efficient factored abstraction for reinforcement learning.
Coskun SahinErkin ÇildenFaruk PolatPublished in: CYBCONF (2015)
Keyphrases
- memory efficient
- reinforcement learning
- state space
- state abstraction
- factored markov decision processes
- temporal abstractions
- external memory
- high level
- function approximation
- reinforcement learning algorithms
- markov decision processes
- model free
- iterative deepening
- robotic control
- learning algorithm
- optimal policy
- decision theoretic planning
- temporal difference
- partially observable
- multiple sequence alignment
- reward function
- path finding
- optimal control
- heuristic search
- multi agent systems
- computer vision