SWIRL: Selection of Workload-aware Indexes using Reinforcement Learning.
Jan KossmannAlexander KastiusRainer SchlosserPublished in: EDBT (2022)
Keyphrases
- reinforcement learning
- response time
- function approximation
- index selection
- state space
- database
- index structure
- multi agent reinforcement learning
- physical design
- temporal difference
- model free
- optimal control
- b tree
- markov decision processes
- query processing
- multi agent
- machine learning
- real world
- neural network
- data sets