| Description |
1 online resource (35 pages) |
| Summary |
Machine learning has accelerated in several industries recently, enabling companies to automate decisions and act based on predicted futures. In time, nearly all major industries will embed ML into the core of their businesses, but right now the gap between companies that successfully adopt ML and those that fail continues to grow. This report examines why so many ML initiatives stall, especially at the stage of moving models from proof of concept to production. Authors Ben Epstein and Paige Roberts examine the strengths and weaknesses of data lake and data warehouse analytic architectures, including the ways that companies use them cooperatively in production. You'll learn how to merge these separate technology stacks into a unified architecture that will streamline the daily workflows of data scientists and data engineers, and facilitate the seamless transition of models from development into production. With this report, you'll explore: Why the unique challenges of MLOps have caused so many ML applications to fail The evolution of data warehouse and data lake architectures How a unified analytics architecture enables you to unite the workflows of business analysts and data scientists How this architecture helps you get new ML projects into production as easily as creating new tables on a dashboard The advantages of in-database ML, including enhanced security, speed and scalability, accessibility, governance, and production readiness. |
| Subject |
Machine learning.
|
|
Apprentissage automatique. |
|
Machine learning |
| Added Author |
Roberts, Paige, author
|
|
More Resources