Termination Approximation: Continuous State Decomposition for Hierarchical Reinforcement Learning.
Sean HarrisBernhard HengstMaurice PagnuccoPublished in: AAAI Workshop: Knowledge, Skill, and Behavior Transfer in Autonomous Robots (2015)
Keyphrases
- continuous state
- hierarchical reinforcement learning
- reinforcement learning
- model free
- state abstraction
- finite state
- state dependent
- function approximation
- robot navigation
- control policies
- action space
- queueing networks
- multi agent
- partially observable markov decision processes
- state action
- state space
- planning problems
- average reward
- dynamic programming
- reward function
- markov decision process
- long run
- hierarchical decomposition
- markov decision processes
- optimal policy