Towards autonomic urban traffic control with collaborative multi-policy reinforcement learning.
Ivana DusparicJulien MonteilVinny CahillPublished in: ITSC (2016)
Keyphrases
- reinforcement learning
- optimal policy
- urban traffic
- urban traffic control
- policy search
- action selection
- markov decision process
- multi agent
- state space
- function approximation
- dynamic programming
- policy gradient
- control policy
- learning algorithm
- traffic control
- action space
- reward function
- markov decision processes
- traffic flow
- reinforcement learning algorithms
- model free
- long run
- real time