Reward Function Optimization of a Deep Reinforcement Learning Collision Avoidance System.
Cooper ConeMichael P. OwenLuis E. AlvarezMarc BrittainPublished in: CoRR (2022)
Keyphrases
- collision avoidance
- reward function
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- path planning
- state space
- mobile robot
- optimal policy
- inverse reinforcement learning
- transition model
- partially observable
- dynamic environments
- markov decision process
- multiple agents
- function approximation
- hierarchical reinforcement learning
- transition probabilities
- model free
- initially unknown
- machine learning
- formation control
- learning algorithm
- temporal difference
- generative model
- dynamic programming
- multi agent
- path finding
- prior knowledge