Chapter 9. Developing Menu Options
I showed you in previous chapters how you can help decision-makers make the best choice from a menu of choice possibilities. This was all predicated on knowing the menu. The menu, however, must be developed somehow and by someone. This simple observation leads to an important question I will address in this chapter: “Where does the menu come from?”
You, as the data scientist, along with the DAC, have a major input into menu development. The decision-makers themselves, however, also contribute to creating it through the questions they ask you and the DAC. This broadens their scope and responsibilities, making them the deciders and, simultaneously, the creators of the very menu they must decide on. The questions are what-if questions reflecting their curiosity about possible, unforeseen events and their impacts.
In this chapter, I will address the following leading questions:
-
What is the difference between what-if and scenario analyses?
-
How can menu items be generated using what-if questions and scenarios?
-
How do you do what-if analyses using a basic and an advanced, more complicated predictive model?
-
How do you generate a synthetic data set for scenario analysis?
I will first introduce in this chapter What-if Analysis for non-stochastic Prescriptive Analytics. This allows you to develop a menu with each option as a different what-if case. A stochastic approach, however, can be used to generate synthetic data for more complicated scenario-type ...
Get Hands-On Prescriptive Analytics now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.