PRO+ Premium Content/Business Information

Thank you for joining!
Access your Pro+ Content below.
February 2017, Vol. 5, No. 1

When predictive analytics models produce false outcomes

Predictions are very difficult, especially when they're about the future. Whether attributed in part or whole to Nobel physicist Niels Bohr or Hall of Fame catcher Yogi Berra, at no time was that prescient observation more applicable than in predicting the outcome of the 2016 presidential election.

The cover story of Business Information's February issue examines some of the hard lessons that, in retrospect, need to be applied when it comes to building and running predictive analytics models, be it in the political realm or the business world. Those lessons include working with high-quality data, measuring the right information, and making sure predictive models reflect reality and not wishful thinking.

In another feature, we look at a critical first step in ensuring that predictive analytics models yield valid outcomes. The proper preparation of data -- a process made more complex with the proliferation of big data and the expansion of predictive analytics applications to leverage it -- requires new tools and techniques.

Also in this issue, learn how the self-proclaimed "CRM and analytics czar" of pro basketball's Charlotte Hornets convinced team ownership to invest in Phizzle FanTracker, which helped combine millions of fan records into accurate, up-to-date, individual fan profiles.

Features in this issue

Columns in this issue

SearchBusinessAnalytics

SearchAWS

SearchContentManagement

SearchOracle

SearchSAP

SearchSQLServer

-ADS BY GOOGLE

Close