SQLFlow: A Flexible Toolkit for Integrating Databases and AI
Authors: Jun Zhou, Ke Zhang, Lin Wang, Hua Wu, Yi Wang, ChaoChao Chen; Published in 2023, 24(116):1−9.
Abstract
The integration of AI algorithms into databases is an ongoing endeavor in both academia and industry. In this paper, we present SQLFlow, a versatile toolkit that seamlessly combines data manipulations and AI operations. SQLFlow allows for local or remote execution and extends SQL syntax to support various AI tasks such as model training, inference, interpretation, and mathematical optimization. It is compatible with multiple database management systems (DBMS) and AI engines, including MySQL, TiDB, MaxCompute, Hive, TensorFlow, scikit-learn, and XGBoost. Detailed documentation and case studies can be found at https://sqlflow.org. The source code and additional information are available at https://github.com/sql-machine-learning/sqlflow.
[Abstract]
[Code]