Online Learning of Weakly Coupled MDP Policies for Load Balancing and Auto Scaling.
S. R. EshwarLucas Lopes FelipeAlexandre Reiffers-MassonDaniel Sadoc MenaschéGugan ThoppePublished in: CoRR (2024)
Keyphrases
- load balancing
- online learning
- optimal policy
- markov decision process
- markov decision processes
- reward function
- dynamic load balancing
- scheduling policies
- state space
- distributed systems
- reinforcement learning
- mobile agents
- active learning
- e learning
- fault tolerance
- dynamic programming
- long run
- infinite horizon
- round robin
- load balance
- grid computing
- fault tolerant
- resource utilization
- peer to peer
- parallel database systems
- skewed data
- load balancing strategy
- artificial intelligence
- peer to peer systems
- low overhead
- data replication
- database systems
- partial replication
- machine learning
- proxy servers
- average cost
- load balancing strategies