Geometric deep reinforcement learning for dynamic DAG scheduling.
Nathan GrinsztajnOlivier BeaumontEmmanuel JeannotPhilippe PreuxPublished in: SSCI (2020)
Keyphrases
- reinforcement learning
- state space
- scheduling problem
- directed acyclic graph
- function approximation
- real time
- resource allocation
- round robin
- deep learning
- dynamically changing
- geometric information
- temporal difference
- model free
- scheduling algorithm
- dynamic environments
- dynamic programming
- lower bound
- data structure
- multi agent
- learning algorithm