GTA: Generative Trajectory Augmentation with Guidance for Offline Reinforcement Learning.
Jaewoo LeeSujin YunTaeyoung YunJinkyoo ParkPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- generative model
- function approximation
- reinforcement learning algorithms
- state space
- data driven
- unsupervised learning
- supervised learning
- temporal difference
- real time
- trajectory data
- multi agent reinforcement learning
- moving object trajectories
- model free
- information management
- robot control
- temporal difference learning
- stochastic approximation
- reward function
- transition model
- robotic control
- learning problems
- markov decision processes
- transfer learning
- dynamic programming
- learning process
- information infrastructure
- spatio temporal
- trajectory planning
- autonomous learning
- multi agent
- learning algorithm
- policy search
- machine learning
- neural network