Improving Scheduling Performance Using a Q-Learning-Based Leasing Policy for Clouds.
Alexander FöllingMatthias HofmannPublished in: Euro-Par (2012)
Keyphrases
- optimal policy
- action selection
- reinforcement learning
- scheduling problem
- learning algorithm
- admission control
- state space
- resource allocation
- stochastic approximation
- learning rate
- multi agent
- scheduling policies
- scheduling algorithm
- cooperative
- markov decision processes
- long run
- markov decision process
- round robin
- actor critic
- model free reinforcement learning
- storage cost
- state action
- reward function
- resource constraints
- evaluation function
- quality of service
- cloud computing
- dynamic programming
- information systems