Interpretable Option Discovery using Deep Q-Learning and Variational Autoencoders.
Per-Arne AndersenOle-Christoffer GranmoMorten GoodwinPublished in: CoRR (2022)
Keyphrases
- image processing
- denoising
- methods in computer vision
- image segmentation
- cooperative
- reinforcement learning
- function approximation
- data mining
- multi agent
- learning algorithm
- knowledge discovery
- state space
- classification rules
- computer vision
- discovery process
- scientific discovery
- deep belief networks
- unsupervised learning
- optimal policy
- model free
- reinforcement learning algorithms
- variational methods
- fixed point
- decision making
- stochastic approximation