Getting Started with Machine Learning in the Cloud

Executive Overview

“Machine learning” is a term that we hear virtually everywhere today. Everyone, it seems, is getting into it, and for good reason. Companies are gaining competitive advantage, delivering better customer experiences, increasing revenue, reacting more swiftly to market shifts—all by applying machine-learning techniques and technologies to the data they already possess.

But challenges abound as well. Data scientists with the necessary know-how are scarce, and they sometimes work in organizational silos rather than in collaboration with other stakeholders in the technical and business groups. It can be difficult to integrate machine-learning models with existing or new applications or processes. And a host of security issues in this area have yet to be resolved.

In this report, we go over the types of real-world applications for machine learning that are delivering successes today. We review the benefits and challenges alike of deploying machine learning. We also explain why most organizations are currently implementing their machine-learning projects in the cloud. And we provide you with a list of best practices that early adopters have found useful in their machine-learning deployments.

Introduction: The Data Opportunity

Think of all the data that is collectively accumulating across different areas of the enterprise. Terabytes and terabytes of it. Just imagine if you could really put this data to work for you. ...