Abstract: The prevailing reinforcement-learning-based traffic signal control methods are typically staging-optimizable or duration-optimizable, depending on the action spaces. In this paper, we use ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback