2019
Learned cardinalities: Estimating correlated joins with deep learning
Publication
Publication
Presented at the
9th Biennial Conference on Innovative Data Systems Research, CIDR 2019 (January 2019), Pacific Grove
We describe a new deep learning approach to cardinality estimation. MSCN is a multi-set convolutional network, tailored to representing relational query plans, that employs set semantics to capture query features and true cardinalities. MSCN builds on sampling-based estimation, addressing its weaknesses when no sampled tuples qualify a predicate, and in capturing join-crossing correlations. Our evaluation of MSCN using a real-world dataset shows that deep learning signiicantly enhances the quality of cardinality estimation, which is the core problem in query optimization.
| Additional Metadata | |
|---|---|
| doi.org/10.48550/arXiv.1809.00677 | |
| 9th Biennial Conference on Innovative Data Systems Research, CIDR 2019 | |
| Organisation | Centrum Wiskunde & Informatica, Amsterdam (CWI), The Netherlands |
|
Kipf, A., Kipf, T., Radke, B., Leis, V., Boncz, P., & Kemper, A. (2019). Learned cardinalities: Estimating correlated joins with deep learning. In CIDR 2019 - 9th Biennial Conference on Innovative Data Systems Research. doi:10.48550/arXiv.1809.00677 |
|
