Name
Abstract | ||
|---|---|---|
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 significantly enhances the quality of cardinality estimation, which is the core problem in query optimization. |
Year | Venue | DocType |
|---|---|---|
2019 | CIDR | Conference |
Volume | Citations | PageRank |
abs/1809.00677 | 6 | 0.40 |
References | Authors | |
25 | 6 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Andreas Kipf | 1 | 32 | 11.03 |
| Thomas N. Kipf | 2 | 796 | 20.67 |
| Bernhard Radke | 3 | 14 | 1.86 |
| Viktor Leis | 4 | 425 | 30.26 |
| Peter Boncz | 5 | 2517 | 244.81 |
| Alfons Kemper | 6 | 3519 | 769.50 |