Interpretable Policy Specification and Synthesis through Natural Language and RL.
Pradyumna TambwekarAndrew SilvaNakul GopalanMatthew C. GombolayPublished in: CoRR (2021)
Keyphrases
- natural language
- optimal policy
- reinforcement learning
- markov decision process
- formal language
- action selection
- control policy
- markov decision processes
- actor critic
- action space
- policy gradient
- policy search
- semantic analysis
- formal languages
- policy iteration
- reinforcement learning problems
- information extraction
- state and action spaces
- decision problems
- control policies
- state space
- partially observable domains
- formal specification
- natural language processing
- learning algorithm
- partially observable markov decision processes
- natural language interface
- machine learning
- specification language
- question answering
- natural language generation
- rl algorithms
- markov decision problems
- high level
- approximate dynamic programming
- reinforcement learning algorithms
- model free reinforcement learning
- function approximators
- knowledge representation
- optimal control
- finite state
- program synthesis
- access control policies
- sequential decision making