Data Smart, 2nd Edition

9Optimization Modeling: Because That “Fresh-Squeezed” Orange Juice Ain't Gonna Blend Itself

Imagine you worked for a large beverage company. You are tasked with creating the perfect blend of orange juices for your not-from-concentrate product.

How would you do this? Clustering? Forecasting? AI?

In fact, you would use none of these. Instead, you would use what's called an optimization model. Huh? What's the difference?

Let's think about this for a second. An artificial intelligence model predicts the result of a process by analyzing its inputs. So far, that's what we've done in this book.

But, here, we're not looking to predict an outcome from data; rather, we need to create the perfect product mix that minimizes costs and maximizes profit. To do this, we'll have to understand things such as inventory, demand, specs, and so on.

We'll see that you can't have it all: you can change the specs for instance, but then you might limit the amount of space you have in your inventory. An optimization model is all about understanding these trade-offs so that you can make the best decision. That is to say, there's an underlying economy to the whole thing.

You can see then how optimization models are so important to businesses. Companies across industries use them every day to answer questions such as these:

• How do I schedule my call center employees to accommodate their vacation requests, balance overtime, and eliminate back-to-back graveyard shifts for any one employee?
• Which oil drilling ...