With Prescriptive Analytics, the future ain't what it used to be

The late great baseball legend Yogi Berra was credited with saying this gem: "The future ain't what it used to be." In the context of big data analytics, I am now inclined to believe that Yogi was very insightful -- his statement is an excellent description of Prescriptive Analytics.

Prescriptive Analytics goes beyond Descriptive and Predictive Analytics in the maturity framework of analytics. "Descriptive" analytics delivers hindsight (telling you what did happen, by generating reports from your databases), and "predictive" delivers foresight (telling you what will happen, through machine learning algorithms). Going one better, "prescriptive" delivers insight: discovering so much about your application domain (from your collection of big data and information resources, through data science and predictive models) that you are now able to take the actions (e.g., set the conditions and parameters) needed to achieve a prescribed (better, optimal, desired) outcome. 

So, if predictive analytics can use historical training data sets to tell us what will happen in the future (e.g., which products a customer will buy; where and when your supply chain will need replenishing; which vehicles in your corporate fleet will need repairs; which machines in your manufacturing plant will need maintenance; or which servers in your data center will fail), then prescriptive analytics can alter that future (i.e., the future ain't what it used to be).

Each of these levels of analytics maturity is accessible to businesses and organizations of all sizes. They are not restricted to large organizations only. In fact, many studies, lessons learned, and data analytics veterans will tell you to start by going after the "low-hanging" fruit. In other words, don't aim too high. It is better to think big (strategically), but start small (tactically). See the excellent article related to this topic "Why do so many analytics projects fail? Key considerations for deep analytics on big data, learning and insights" and also check out my article "Machine Unlearning: The Value of Imperfect Models."

With so much attention given to the complexity of big data collections and the systems that manage and process them, we can easily overlook the fact that sometimes the fastest path to a solution (perhaps one innovation, or one monetized data product) is the simplest thing: the MVP ("Minimum Viable Product"). By achieving easy-wins in the short term, data analytics teams can demonstrate the ROI (Return On Innovation) of their analyses to their broader organization, thereby ensuring ongoing support for long-term analytics goals.

I have devised a simple way to mathematically illustrate and distinguish predictive vs. prescriptive analytics:

• Predictive: Given X, find Y.
• Prescriptive: Given Y, find X.

The following graphic illustrates the concept:

When dealing with large high-variety data sets, with many features and measured attributes, it is often difficult to build accurate models that are generally useful under a variety of conditions and that capture all of the complexities of the response functions and explanatory variables within your business application. In such cases, fast automatic modeling tools are needed. These tools can help to identify the minimum viable feature set for accurate predictive and prescriptive modeling. Those capabilities are the "secret sauce" in insightful prescriptive analytics, and they coincide nicely with another insightful quote from Yogi Berra: "You can observe a lot by just watching."

Follow Kirk Borne on Twitter @KirkDBorne