Enhanced Query Featurization for Cardinality Estimation

Updated implementation of multi-set convolutional networks (MSCN) to include and test other featurizations [1].

Requirements

• PyTorch 1.0
• Python 3.7

Usage

python3 train.py --help

Example usage:

python3 train.py synthetic

To use a different featurization:

python3 train.py --feat range --queries 100000 --epochs 100 synthetic

You can find the different featurizations in mscn\util.py.

References

[1] Müller et al., Enhanced Featurization of Queries with Mixed Combinations of Predicates for ML-based Cardinality Estimation , 2023

Cite

Please cite our paper if you use this code in your own work:

@inproceedings{mueller2023enhanced, 
  doi = {10.48786/EDBT.2023.22}, 
  url = {https://openproceedings.org/2023/conf/edbt/paper-1.pdf}, 
  author = {M\"uller, Magnus and Woltmann, Lucas and Lehner, Wolfgang}, 
  keywords = {Database Technology}, 
  language = {en}, 
  title = {{Enhanced Featurization of Queries with Mixed Combinations of Predicates for ML-based Cardinality Estimation}}, 
  publisher = {OpenProceedings.org}, 
  year = {2023}, 
  booktitle = {Proceedings of the 26th International Conference on Extending Database Technology}, 
  location = {Ioannina, Greece},
  series = {EDBT 2023}}