Policy Search on Aggregated State Space for Active Sampling.
Sandeep ManjannaHerke van HoofGregory DudekPublished in: ISER (2018)
Keyphrases
- policy search
- active sampling
- state space
- reinforcement learning algorithms
- reinforcement learning
- reward function
- dynamic programming
- selective sampling
- active learning
- markov decision problems
- partially observable markov decision processes
- continuous state
- learning to rank
- free form
- markov decision processes
- heuristic search
- optimal policy
- markov chain
- action space
- feature selection
- model free
- dynamical systems
- state variables
- temporal difference
- machine learning
- partially observable
- random sampling
- ranking functions
- particle filter
- function approximation
- planning problems
- belief state
- function approximators
- supervised learning
- learning process
- policy gradient
- learning algorithm
- decision theoretic
- stochastic processes
- search space