Book Image

Analytics: How to Win with Intelligence

By : John Thompson, Shawn P. Rogers
Book Image

Analytics: How to Win with Intelligence

By: John Thompson, Shawn P. Rogers

Overview of this book

Today, business is moving into an era where information is more valuable than services. Organizations that connect information with their products will have a huge advantage. This book helps people understand the power of data analytics and explains how some of the tools available can be applied to a wide range of applications. It begins with a brief history of analytics and explains how it all began. You'll learn about several common analytical approaches and the tools that data scientists use to analyze data. You'll gain insight into some staffing models, technologies, organizational structures, and analytical approaches used in the previous two eras of analytics. As you progress through the chapters, you'll also get a glimpse into the future of the analytical marketplace. After reading this book, you will be able to help your team deploy analytical elements into your operations and become competitive in your business.
Table of Contents (11 chapters)
Free Chapter
1
Foreword by Tom Davenport

The process of advanced analytics

There are three general stages of the advanced analytics process: model building, scoring, and maintenance. The following is a brief summary of each of those steps.

Once a company has determined the business objective of the project, the model building process begins. This is where the analytics team examines the data, determines which analytical techniques are appropriate, and then begins the iterative process of creating new models.

Typically, the models are built, tested, and improved through iterative cycles. Once the model is performing at a level at which the analytical team and the business team are happy with the results, the process is ready to move to the next step.

We’re often asked how long the model-building process typically takes. That’s an excellent question that’s unfortunately very difficult to answer. In the past, the process could take years to get a model to the point at which it’s ready...