Db2 AI Query Optimizer: Expanded Predicate Support and Improved Model Training
In our previous blog posts (1, 2, 3, 4), we introduced the Db2 AI Query Optimizer, a revolutionary approach to database query optimization using neural networks. We demonstrated how our system learns from table column correlations to make more intelligent decisions about query execution plans, outperforming traditional optimizers in many scenarios.
Today, we're excited to share 2 significant enhancements to the Db2 AI Query Optimizer that further improve its performance and capabilities:
- Leveraging of Transfer Learning to reduce model training time and improve model robustness
- Expanded support for the 'IS NULL' predicate
These new features represent our ongoing commitment to making database query optimization smarter, faster, and more efficient.
The best part? You don't need to do anything to take advantage of these improvements. If you are already using the Db2 AI Query Optimizer, these improvements are automatically enabled as soon as you upgrade to Db2 version 12.1.2 or later.
IDUG Members have access to premium content – better yet, membership is FREE. Click here to create your member profile today. Questions? Contact us!