Sample-efficient Reinforcement Learning Representation Learning with Curiosity Contrastive Forward Dynamics Model.
Thanh NguyenTung Minh LuuThang VuChang D. YooPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- learning process
- dynamic bayesian networks
- computational model
- learning scheme
- bayesian networks
- learning problems
- experimental data
- supervised learning
- hidden state
- learning algorithm
- mathematical model
- learning phase
- probabilistic model
- prior knowledge
- high level
- adaptation process
- autonomous learning
- machine learning
- qualitative models
- transition model
- learning models
- learning tasks
- sample size
- learning systems
- em algorithm
- markov random field
- active learning