A Paradigm for Assessing the Scope and Performance of Predictive Analytics
28 Pages Posted: 12 Jul 2018 Last revised: 15 Oct 2018
Date Written: June 20, 2018
Abstract
In this paper, I outline possibilities and limitations for the scope and performance of predictive analytics within a simple paradigm. I do this by first bifurcating predictive analytics into two categories, passive and active. I contrast this categorization with current alternatives and highlight its relative merits in terms of clarity in boundaries, as well as appropriate methods for different types of prediction. I then describe the range of suitable applications, as well as the possibilities and limitations with regard to prediction accuracy, for each type of prediction. I conclude with a discussion of key ways in which an understanding of this paradigm can be valuable.
Keywords: Predictive Analytics, Paradigm, Framework, Prediction, Artificial Intelligence, Machine Learning, Econometrics, Data Mining, Diagnostics
JEL Classification: C1
Suggested Citation: Suggested Citation