Generalization in Deep Reinforcement Learning for Robotic Navigation by Reward Shaping.
Victor R. F. MirandaArmando Alves NetoGustavo Medeiros FreitasLeonardo A. MozelliPublished in: IEEE Trans. Ind. Electron. (2024)
Keyphrases
- reward shaping
- robotic navigation
- reinforcement learning
- complex domains
- reinforcement learning algorithms
- function approximation
- markov decision problems
- multi agent
- state space
- supervised learning
- model free
- reward function
- dynamic programming
- learning algorithm
- policy iteration
- machine learning
- learning capabilities
- partially observable
- markov decision processes
- domain knowledge