Designing an offline reinforcement learning objective from scratch.
Gaon AnJunhyeok LeeXingdong ZuoNorio KosakaKyung-Min KimHyun Oh SongPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- learning algorithm
- function approximation
- real time
- multi agent
- state space
- model free
- multiple objectives
- optimal control
- expert systems
- multi objective
- hidden markov models
- database
- support vector
- multiscale
- search engine
- real world
- action selection
- temporal difference
- reinforcement learning algorithms
- learning capabilities
- partially observable
- robot control
- markov decision process
- temporal difference learning