An example outcome of descriptive models is categorizing customers by their
product preferences and age. Example methodologies for this categorization are
classification and clustering. We will learn more about it in the following sections. As
a next step to this categorization, descriptive models can help model a large number
of individual elements and help make predictions.
Predictive analytics is all about turning data into valuable and actionable information.
Predictive analytics employs all the attributes of data analyzed as a part of descript-
ive analytics to conclude a probable future outcome, given a situation context.
Predictive analytics paradigm includes running a variety of statistical techniques to
analyze historical and current behavior to make predictions about future events.
Predictive models can identify and exploit specific customer patterns to determine
possiblerisksandopportunities, givenaparticularconditionorcontext. Thefollowing
are the three important aspects of predictive analytics:
• Decision analysis
• Optimization and profiling
Let us look at some examples where predictive analytics can be used:
• Customer relationship management systems : Using predictive analytic
techniques, we can analyze the entire customer data, identify patterns, and
predict their behavior.
• Product management and sales : For a company that offers multiple
products, we can use predictive analytics to analyze the spending patterns
of customers and identify cross sales or additional sales, thus paving way to
higher profitability. This is also referred to as cross-selling or up-selling.
It is important that a strong team of business or domain experts and data scientists
is formed who understands statistical modeling techniques and can apply these on
data to derive business insights. The problems that we are solving and the questions
we are answering are based on the needs of the business, which typically are not of
statisticians or data scientists.