Data Management and Data Warehouse Requirements for Machine Learning and AI
Webinar Speaker: Philip Russom, Senior Research Director for Data Management
Date: Wednesday, November 20, 2019
Time: 9:00 a.m. PT, 12:00 p.m. ET
Getting the right data, in the right form, on the right platform, for the right analytics method.
In this TDWI webinar, speakers will begin by defining machine learning (ML), plus related technologies and practices in artificial intelligence and predictive analytics. The presentation will also discuss ML’s compelling use cases in business analytics and the enablement of automated responses that do not require human intervention. The use cases covered will include: sales recommendations, offers that halt customer churn, automatic personalization of marketing campaigns, fraud prediction, preventative maintenance schedules for machinery, and trip routing.
However, the core of the webinar will help explain the many data management requirements that users must address in order to successfully apply machine learning to their analytics. In particular, attendees will learn about the typical ML lifecycle stages and the unique data requirements of each:
- Understand business problems and the available data to solve for them
- Discover data for initial insights and solution definition
- Obtain exploratory data for a data scientist or other technical user to manually develop an analytic model
- Apply training data so a smart algorithm can automatically generate the model
- Use production data to drive continuous improvement and address model drift, whether revisions of the model are automatic or developer guided
Satisfying ML’s diverse data requirements involves a modern data management infrastructure, typically including cloud-based platforms for data warehousing, data lakes, data integration, and more.
Philip Russom, Ph.D.