From Two-Dimensional to Three-Dimensional Environment with Q-Learning: Modeling Autonomous Navigation with Reinforcement Learning and no Libraries.
Ergon Cugler de Moraes SilvaPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- three dimensional
- autonomous navigation
- mobile robot
- dynamic environments
- reinforcement learning algorithms
- function approximation
- real robot
- multi agent
- model free
- action selection
- state space
- outdoor environments
- learning agent
- reinforcement learning methods
- optimal policy
- topological map
- path planning
- temporal difference
- stochastic approximation
- exploration strategy
- surface reconstruction
- d objects
- learning algorithm
- machine learning
- ego motion
- markov decision processes
- dynamic programming
- multiscale
- real time
- depth map
- image sequences
- agent learns