Continuous-Action Reinforcement Learning for Memory Allocation in Virtualized Servers.
Luis A. GarridoRajiv NishtalaPaul M. CarpenterPublished in: ISC Workshops (2019)
Keyphrases
- np hard
- continuous action
- policy search
- reinforcement learning
- data center
- continuous state
- action space
- partially observable markov decision processes
- cloud computing
- continuous state and action spaces
- reinforcement learning algorithms
- state space
- function approximation
- model free
- decision problems
- markov decision processes
- dynamic programming
- robot navigation
- reward function
- finite state
- control policies
- state dependent
- multi agent
- hidden state
- function approximators
- learning algorithm
- optimal control
- stochastic processes
- partially observable
- reinforcement learning methods
- arrival rate
- action selection
- policy gradient
- evaluation function
- transfer learning
- machine learning