Hands-On Prescriptive Analytics

Chapter 2. Prescriptive Methods: Overview

The many Prescriptive Analytics methodologies available for narrowing a decision menu can easily overwhelm and stymie your analytical work. It is helpful, therefore, for you to not only know them, but also to know their pros and cons so you can efficiently and effectively work with them. Fortunately, only a few, I believe, are important. Therefore, the leading questions for this chapter are:

• What are the primary methods mentioned in the Prescriptive Analytics literature?

• How can we categorize the methods?

• Can we create general umbrella categories to aid understanding?

• What are the most important methods?

The umbrella categories are non-stochastic and stochastic methods. They form the structure of Parts III and IV of this book.

Introduction to Prescriptive Analytics Methods

Trying to name, much less summarize, the many Prescriptive Analytics methodologies is a challenge, to say the least. A simple first, but not useful, approach is to group them by classes. There are two major classes:

• Proprietary

• Public domain

Proprietary Methods

Proprietary methods are the intellectual property of consulting firms and the internal data science divisions of major enterprises. In each instance, there is a definite competitive advantage from not disclosing how decisions are made, including the methods used to develop and provide rich information. This is not an issue for governmental agencies that are transparent in this regard because ...