Zero-shot Learning of Hint Policy via Reinforcement Learning and Program Synthesis.
Aleksandr EfremovAhana GhoshAdish SinglaPublished in: EDM (2020)
Keyphrases
- program synthesis
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- action selection
- partially observable environments
- reward function
- state and action spaces
- reinforcement learning problems
- function approximators
- partially observable
- actor critic
- policy gradient
- function approximation
- policy evaluation
- control policy
- reinforcement learning algorithms
- recursive programs
- state space
- knowledge transfer
- markov decision processes
- transfer learning
- temporal difference
- inductive logic programming
- action space
- rl algorithms
- learning algorithm
- markov decision problems
- average reward
- partially observable markov decision processes
- state action
- optimal control
- machine learning
- policy iteration
- databases
- model free
- inverse reinforcement learning
- supervised learning
- agent learns
- dynamic programming
- learning process