Human-like Car Following

Deep Reinforcement Learning; Deterministic Policy Gradient

Project:

Using Deep Reinforcement Learning to design a Human-like Car-Following system

  • Leveraged an Actor-critic Reinforcement Learning framework to learn personalized car-following behaviors from empirical data
  • Employed Deep Deterministic Policy Gradient (DDPG) to learn the continuous-control policy network.
  • MDP
  • Bandit
  • Actor-critic Reinforcement Learning
  • Deep deterministic policy gradient

Meta Information:

  • Time: Fall 2020
  • Instructor: Hongning Wang
  • Affiliation: University of Virginia
  • Teamate: Xingbo Fu