Reinforcement Learning Applicability for Resource-Based Auto-scaling in Serverless Edge Applications.
Priscilla BenedettiMauro FemminellaGianluca RealiKris SteenhautPublished in: PerCom Workshops (2022)
Keyphrases
- reinforcement learning
- edge detection
- state space
- real world
- function approximation
- machine learning
- resource allocation
- dynamic programming
- optimal policy
- edge information
- resource management
- supervised learning
- learning process
- markov decision processes
- learning algorithm
- optimal control
- agent receives
- edge direction
- temporal difference
- web resources
- multi agent
- weighted graph
- resource constraints
- model free
- edge detector
- reinforcement learning algorithms
- action space
- resource consumption
- stochastic approximation
- multi agent reinforcement learning
- scale space