Computer Science
2 years ago

Revolutionizing Data Systems: Query Superoptimization to Accelerate Queries

arXiv (Cornell University)
Learned Query Superoptimization
Ryan Marcus

Paper Summary

Related papers

Theme background image
Computer Science
8 years ago

Streaming Data Queries Revolutionize Real-Time Decision-Making Across Industries

Data Stream Management
Semantic Web
Trajectory Analysis
4 views
Theme background image
Computer Science
a year ago

Database Query Forms Revolutionize User Experience, Empowering Seamless Information Access

Trajectory Analysis
Data Stream Management
Data Mining Techniques
4 views
Theme background image
Computer Science
2 years ago

Revolutionizing Data Systems: Query Superoptimization to Accelerate Queries

Active Learning
Robotics Coordination
Multi-Armed Bandit
1 view

Query optimizers can learn from mistakes to improve query plans, but what if they could avoid mistakes altogether? A new concept called learned query superoptimization suggests that query optimizers can autonomously experiment to find the best plans using advanced algorithms. While this process may take longer, it has the potential to greatly speed up repetitive queries in data systems.