- 138 views
- 195 downloads
A Neural Network for Cardinality Estimation
- Author / Creator
- Xingchi wang
With the booming development of the Internet, the amount of data stored in databases are enormously growing. Databases retrieve relevant information in response to users’ queries; the retrieved information is encoded in dynamically generated by databases in the form of structured data records. As the amount of data increases, the query time becomes longer and longer. So, there is an urgent need for database optimization.
Database optimization is the strategy of reducing database system response time. Databases provide us with information stored with a hierarchical and related structure, which allows us to extract the content and arrange it easily. Database optimization includes avoiding unused tables, using proper indexing, avoiding temporary tables and coding loops and so on. In our research, we choose to optimize cardinality estimation in database optimizer.
Cardinality estimation is a fundamental task in database query processing and optimization. However, the accuracy of traditional estimation techniques is poor resulting in non-efficient query execution plans. With the rise of deep learning, there is a general notion that data representation can lead to better estimation accuracy. Up to now, all proposed neural network approaches for cardinality estimation can only deal with inner joins between tables. To overcome this issue, we introduce a novel neural network (NN) in this paper. Through systematic experiments and scientific analysis results, it is proved that our model performs better than other models. This approach leads to better data representation and thus better estimation accuracy in multiple types of joins.
- Graduation date
- Spring 2021
- Type of Item
- Master of Science
- This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.