Embedding AI/ML in applications using OML in Autonomous Database
In this Autonomous Database Learning Lounge session, Oracle Machine Learning and Autonomous Database Product Management introduced Oracle Machine Learning in-database capabilities, as well as Python, R, REST and the latest no-code UI options for this component that is included with every Autonomous Database.
Slides at: https://bit.ly/adbll-slides-oml
Outline
Introduction to Oracle Machine Learning concepts and capabilities in ADB
- Mark Hornick, Senior Director of Product Management introduced concepts surrounding in-database machine learning from SQL, Python, R and REST APIs, along with OML components exclusive to Autonomous Database. He highlighted the benefits of using machine learning in ADB, common use cases, and the upcoming features.
Oracle Machine Learning UI for Autonomous Database
- Marcos Arancibia, Senior Principal Product Manager will showed a quick live demo of the different UI options available for using OML with Autonomous Database and the no-code OML AutoML UI.
Oracle Machine Learning for Developers: Python and R Third-Party packages, REST APIs and Model Monitoring
- Sherry LaMonica, Consulting Member of Technical Staff showed how to use Conda environments for third-party Python and R packages with Autonomous Database and how to use ML models deployed to OML Services using the REST API for real-time scoring, followed by invoking user-defined Python and R functions through REST and SQL APIs, and the use of Model Monitoring for detecting the need to replace an ML model in production.
Open Q&A
- Autonomous Database product managers answered technical questions about the Autonomous Database service, its features, integration capabilities and use cases.
Video highlights:
00:43 Agenda
02:03 Important links on Autonomous Database
02:24 Introduction to Oracle Machine Learning
03:50 Example of use cases and ML Techniques
06:38 Sources of ML models
08:27 Oracle Machine Learning in-Database algorithms
09:12 Build in-database models from OML APIs: SQL, R and Python
10:10 Machine Learning Concepts: Regression
12:15 Machine Learning Concepts: Classification
13:23 Machine Learning process
14:28 Automated Machine learning: Python API and OML AutoML UI
16:08 Using ML models in your applications
16:54 Live Demo: OML AutoML UI and OML Notebooks EA
37:14 Simplify Python and R solution deployment
39:37 Deploy Python and R native models to use in applications
40:00 Customer third-party packages on ADB via OML Notebooks
41:06 Live Demo: Third-party packages plus UDFs via Python/SQL/REST
48:08 Use ML models from REST endpoints with OML Services
49:26 Deploy in-database models to use in applications
50:04 Deploy ONNX-format models for use by applications
50:50 OML Services Data and Model Monitoring
52:50 Live Demo: OML Services - scoring and monitoring
59:21 OML family of components
59:37 OML enhancements for Oracle Database 23c
01:02:09 Autonomous Database as a platform for Data Science and Machine Learning
01:03:55 Important links for you to bookmark
01:04:19 Final thoughts