32.1 Highlights of the Oracle Machine Learning for SQL API

Learn about the advantages of OML4SQL application programming interface (API).

Machine learning is a valuable technology in many application domains. It has become increasingly indispensable in the private sector as a tool for optimizing operations and maintaining a competitive edge. Machine learning also has critical applications in the public sector and in scientific research. However, the complexities of machine learning application development and the complexities inherent in managing and securing large stores of data can limit the adoption of machine learning technology.

OML4SQL is uniquely suited to addressing these challenges. The machine learning engine is implemented in the database kernel, and the robust administrative features of Oracle Database are available for managing and securing the data. While supporting a full range of machine learning algorithms and procedures, the API also has features that simplify the development of machine learning applications.

The OML4SQL API consists of extensions to Oracle SQL, the native language of the database. The API offers the following advantages:

• Scoring in the context of SQL queries. Scoring can be performed dynamically or by applying machine learning models.

• Automatic Data Preparation (ADP) and embedded transformations.

• Model transparency. Algorithm-specific queries return details about the attributes that were used to create the model.

• Scoring transparency. Details about the prediction, clustering, or feature extraction operation can be returned with the score.

• Simple routines for predictive analytics.

• A workflow-based graphical user interface (GUI) within Oracle SQL Developer. You can download SQL Developer free of charge from the following site:

  Oracle Data Miner

Note:

The examples in this publication are taken from the OML4SQL examples that are available on GitHub. For information on the examples, see Oracle Machine Learning for SQL Examples.