End-to-end Reinforcement Learning for Autonomous Longitudinal Control Using Advantage Actor Critic with Temporal Context.
Sampo KuuttiRichard BowdenHarita JoshiRobert de TempleSaber FallahPublished in: ITSC (2019)
Keyphrases
- end to end
- actor critic
- reinforcement learning
- temporal context
- optimal control
- control problems
- approximate dynamic programming
- temporal difference
- policy gradient
- function approximation
- control policy
- reinforcement learning algorithms
- dynamic programming
- neuro fuzzy
- gradient method
- control strategy
- temporal information
- spatial context
- state space
- control system
- spatio temporal
- adaptive control
- policy iteration
- audio visual
- markov decision processes
- action selection
- average reward
- computational complexity
- learning algorithm
- infinite horizon
- evaluation function
- multi agent
- optimal policy