Safe Exploration Method for Reinforcement Learning Under Existence of Disturbance.
Yoshihiro OkawaTomotake SasakiHitoshi YanamiToru NamerikawaPublished in: ECML/PKDD (4) (2022)
Keyphrases
- reinforcement learning
- detection method
- high precision
- experimental evaluation
- significant improvement
- cost function
- dynamic programming
- high accuracy
- computationally efficient
- synthetic data
- state space
- clustering method
- prior knowledge
- multiresolution
- probabilistic model
- denoising
- feature set
- pairwise
- segmentation method
- data sets