Explaining a Deep Reinforcement Learning (DRL)-Based Automated Driving Agent in Highway Simulations.
Francesco BellottiLuca LazzaroniAlessio CapelloMarianna CossuAlessandro De GloriaRiccardo BertaPublished in: IEEE Access (2023)
Keyphrases
- reinforcement learning
- multi agent
- action selection
- learning agent
- multi agent systems
- multiagent systems
- state abstraction
- learning capabilities
- autonomous learning
- state action
- agent receives
- markov decision processes
- reward function
- agent model
- autonomous agents
- learning agents
- function approximation
- agent learns
- multiple agents
- reinforcement learning agents
- semi automated
- multi agent reinforcement learning
- exploration strategy
- intelligent agents
- learning algorithm
- traffic accidents
- markov decision process
- partially observable
- fully automated
- reward shaping
- single agent
- machine learning
- multi agent environments
- state space
- action space
- software agents
- neural network
- mobile agents
- model free
- dynamic programming
- real robot
- agent systems
- simulation models
- agent technology
- simulation model
- robocup soccer
- evaluation function