Bansor: Improving Tensor Program Auto-Scheduling with Bandit Based Reinforcement Learning.
Chao GaoTong MoTaylor ZowtukTanvir SajedLaiyuan GongHanxuan ChenShangling JuiWei LuPublished in: ICTAI (2021)
Keyphrases
- reinforcement learning
- scheduling problem
- scheduling algorithm
- higher order
- function approximation
- markov decision processes
- resource allocation
- learning algorithm
- high order
- multi agent
- optimal policy
- state space
- optimal control
- action space
- diffusion tensor
- reinforcement learning algorithms
- temporal difference
- model free
- computer programs
- parallel machines
- resource constraints
- learning classifier systems
- response time
- genetic algorithm
- learning process
- dynamic programming
- transfer learning
- medical images