Danger-aware Weighted Advantage Composition of Deep Reinforcement Learning for Robot Navigation.
Wei ZhangYunfeng ZhangNing LiuPublished in: CoRR (2018)
Keyphrases
- robot navigation
- reinforcement learning
- continuous state
- autonomous mobile robot
- autonomous robots
- scene understanding
- landmark recognition
- real time stereo
- map building
- function approximation
- state space
- initially unknown
- topological map
- optimal policy
- partially observable
- feature vectors
- learning algorithm
- markov decision processes
- markov decision process
- image classification
- feature selection