Experimental design for MRI by greedy policy search.
Tim BakkerHerke van HoofMax WellingPublished in: NeurIPS (2020)
Keyphrases
- experimental design
- policy search
- dynamic programming
- reinforcement learning
- feature selection
- active learning
- reinforcement learning algorithms
- continuous state
- empirical studies
- search algorithm
- sample size
- policy gradient
- reward function
- partially observable markov decision processes
- function approximation
- class imbalance
- multiple agents
- markov decision problems
- state dependent
- principal component analysis
- search space
- np hard