Learned Database Systems: AI-Driven Query Optimization, Learned Indexes, and Cardinality Estimation

Status: public · Confidence: medium (0.8) · Basis: verified_sources

## TL;DR
Learned Database Systems: AI-Driven Query Optimization, Learned Indexes, and Cardinality Estimation: Learned database systems use machine learning to replace or augment database components such as indexes, optimizers, and cost models.

## Core Explanation
The central idea is that data distributions and workloads can sometimes be modeled. Learned components may improve performance, but database systems still require correctness, robustness, update handling, and fallback behavior.

## Further Reading

- [The Case for Learned Index Structures](https://arxiv.org/abs/1712.01208)
- [SageDB: A Learned Database System](https://cidrdb.org/cidr2019/papers/p117-kraska-cidr19.pdf)
- [Bao: Making Learned Query Optimization Practical](https://arxiv.org/abs/2004.03814)