Hindsight Expectation Maximization for Goal-conditioned Reinforcement Learning.
Yunhao TangAlp KucukelbirPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- expectation maximization
- em algorithm
- function approximation
- action selection
- optimal control
- mixture model
- machine learning
- state space
- generative model
- data sets
- parameter estimation
- mobile robot
- gaussian mixture model
- markov decision processes
- maximum a posteriori
- multi agent
- image segmentation
- maximum likelihood estimation
- real time
- autonomous learning