To Support Machine Learning by Database System

J. Yu

In the big data era, machine learning techniques have been extensively studied to learn new things from a huge amount of data, instead of find new things by programming. Given the goal of machine learning is to learn from data, it becomes a natural question how machine learning and database system can be integrated tightly in the same platforms, instead of simply extracting massive data from a database system to conduct machine learning tasks every time when there is such a need, which is with high cost. We concentrate ourselves on supporting machine learning in the kernel of a database system. We focus on query processing techniques, and aim at enhancing query processing to efficiently support machine learning algorithms in a standalone/distributed database system.