Decomposition of Reinforcement Learning for Admission Control of Self-Similar Call Arrival Processes.
Jakob CarlströmPublished in: NIPS (2000)
Keyphrases
- admission control
- reinforcement learning
- control policy
- arrival processes
- end to end
- atm networks
- quality of service
- queue length
- production system
- web server
- resource management
- resource consumption
- state space
- markov decision processes
- learning algorithm
- multi agent
- long run
- optimal policy
- single server
- linear programming
- response time
- dynamic programming
- search space
- web pages