Towards Designing Cost-Optimal Policies to Utilize IaaS Clouds with Online Learning.
Xiaohu WuPatrick LoiseauEsa HyytiäPublished in: ICCAC (2017)
Keyphrases
- optimal policy
- online learning
- average cost
- cloud computing
- markov decision processes
- decision problems
- reinforcement learning
- long run
- lost sales
- state space
- finite state
- finite horizon
- dynamic programming
- infinite horizon
- total cost
- multistage
- state dependent
- expected cost
- policy iteration
- average reward
- average reward reinforcement learning
- markov decision process
- serial inventory systems
- sufficient conditions
- e learning
- initial state
- linear program
- long run average cost
- linear programming
- supply chain
- finite number
- fixed cost
- learning algorithm
- semi markov decision processes
- total reward
- inventory models
- order quantity
- control policies
- active learning
- data management
- cost sensitive