Guiding Safe Reinforcement Learning Policies Using Structured Language Constraints.
Bharat PrakashNicholas R. WaytowichAshwinkumar GanesanTim OatesTinoosh MohseninPublished in: SafeAI@AAAI (2020)
Keyphrases
- reinforcement learning
- optimal policy
- programming language
- policy search
- language learning
- real world
- function approximation
- constraint language
- control policies
- state space
- markov decision processes
- constraint satisfaction
- search space
- neural network
- structured data
- object oriented
- learning process
- constrained optimization
- modeling language
- reward function
- linear constraints
- natural language
- learning algorithm
- fitted q iteration