Analytics off the shelf: Using commercially available tools now

Abstract

Getting started in predictive analytics is arguably quite difficult. The tools used are not the standard statistical tools taught in most universities today. This article is not a course in predictive analytics, but it is a path through which to begin learning to use the tools effectively. The example of a firm attempting to reduce their churn rate is used as a medium for exploring one commonly used predictive analytics algorithm and its most commonly used diagnostic test statistic. A CART model is used to make predictions in two different manners: first, with purely numerical data, and secondly with added text data. The use of social media as a source of information to marketers has become quite useful and this example give the reader some insight as to why this is so.