Depth and nonlinearity induce implicit exploration for RL.
Justas DauparasRyota TomiokaKatja HofmannPublished in: CoRR (2018)
Keyphrases
- reinforcement learning
- exploration exploitation
- exploration strategy
- autonomous learning
- action selection
- depth map
- exploration exploitation tradeoff
- active learning
- neural network
- learning algorithm
- data mining
- state space
- stereo matching
- depth images
- model free
- learning process
- multi agent
- case study
- active exploration