Enhancing Loop-Invariant Synthesis via Reinforcement Learning.
Takeshi TsukadaHiroshi UnnoTaro SekiyamaKohei SuenagaPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- temporal difference
- multi agent
- model free
- affine transformation
- program synthesis
- machine learning
- feedback loop
- action selection
- affine invariant
- optimal policy
- state space
- dynamic programming
- learning process
- learning algorithm
- texture synthesis
- invariant features
- case study
- stochastic approximation
- reinforcement learning algorithms
- learning capabilities
- temporal difference learning
- functional programs
- relational reinforcement learning
- policy search
- database
- learning agents
- robot control
- information systems
- object recognition
- search space
- markov decision processes
- transfer learning