Frontier versus ordinary regression models for data mining

Authors:
Marvin D. Troutt;Michael Hu;Murali Shanker;William Acar
Affiliations:
Kent State University;Kent State University;Kent State University;Kent State University
Venue:
Managing data mining technologies in organizations
Year:
2003

Citing 0
Cited 2

Maximum performance efficiency approaches for estimating best practice costs

Data mining
Data mining for human resource information systems

Data mining

Quantified Score

Hi-index	0.01

Visualization

Abstract

Frontier Regression Models seek to explain boundary, frontier or optimal behavior rather than average behavior as in ordinary regression models. Ordinary regression is one of the most important tools for data mining. Frontier models may be desirable alternatives in many circumstances. In this chapter, we discuss frontier regression models and compare their interpretations to ordinary regression models. Occasional contact with stochastic frontier estimation models is also made, but we concentrate primarily on pure ceiling or floor frontier models. We also propose some guidelines for when to choose between them.