Maximum Entropy Reinforcement Learning with Mixture Policies.
Nir BaramGuy TennenholtzShie MannorPublished in: CoRR (2021)
Keyphrases
- maximum entropy
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- maximum entropy principle
- markov models
- partially observable markov decision processes
- random fields
- reward function
- mixture model
- fitted q iteration
- maximum entropy model
- conditional random fields
- markov decision processes
- principle of maximum entropy
- temporal difference
- transformation based learning
- minimum cross entropy
- model free
- state space
- computer vision
- learning algorithm
- learning problems
- iterative scaling
- class conditional
- machine learning
- exponential family
- bregman divergences
- gaussian mixture model
- unsupervised learning
- active learning