Reinforcement Learning based Control of Imitative Policies for Near-Accident Driving.
Zhangjie CaoErdem BiyikWoodrow Z. WangAllan RaventosAdrien GaidonGuy RosmanDorsa SadighPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- control policy
- control policies
- optimal policy
- control problems
- traffic accidents
- optimal control
- robot control
- policy search
- state space
- markov decision process
- dynamic programming
- decentralized control
- continuous state
- control system
- machine learning
- reinforcement learning agents
- action selection
- model free
- long run
- function approximation
- reward function
- control strategy
- average reward
- inverted pendulum
- transfer learning
- supervised learning
- traffic safety
- mobile robot
- learning process
- multi agent systems