Implementing Deep Reinforcement Learning (DRL)-based Driving Styles for Non-Player Vehicles.
Luca FornerisAlessandro PighettiLuca LazzaroniFrancesco BellottiAlessio CapelloMarianna CossuRiccardo BertaPublished in: Int. J. Serious Games (2023)
Keyphrases
- reinforcement learning
- driving simulator
- dangerous situations
- function approximation
- traffic accidents
- driver assistance systems
- car navigation
- real time
- robotic control
- fuel consumption
- reinforcement learning algorithms
- traffic safety
- intelligent vehicles
- traffic conditions
- vehicle detection
- state space
- road traffic
- imperfect information
- model free
- two player games
- supervised learning
- markov decision processes
- game play
- travel time
- temporal difference
- temporal difference learning
- optimal control
- transfer learning
- deep learning
- learning process
- moving objects
- multi agent
- machine learning