Joint Inference of Reward Machines and Policies for Reinforcement Learning.
Zhe XuIvan GavranYousef AhmadRupak MajumdarDaniel NeiderUfuk TopcuBo WuPublished in: CoRR (2019)
Keyphrases
- reinforcement learning
- joint inference
- reward function
- optimal policy
- total reward
- entity resolution
- information extraction
- probabilistic model
- markov decision process
- reinforcement learning algorithms
- semantic role labeling
- average reward
- conditional random fields
- markov decision processes
- partially observable markov decision processes
- natural language understanding
- graphical models
- control policy
- state space
- dynamic programming
- joint segmentation
- learning algorithm
- supervised learning
- natural language
- exact inference
- variational inference
- record linkage
- data integration
- generative model