A Review of Supervised Classification based on Contrast Patterns: Applications, Trends, and Challenges

Octavio Loyola-González; Miguel Angel Medina-Pérez; Kim-Kwang Raymond Choo

doi:10.1007/s10723-020-09526-y

A Review of Supervised Classification based on Contrast Patterns: Applications, Trends, and Challenges

J Grid Comput. 2020;18(4):797-845. doi: 10.1007/s10723-020-09526-y. Epub 2020 Oct 4.

Authors

Octavio Loyola-González¹, Miguel Angel Medina-Pérez², Kim-Kwang Raymond Choo³

Affiliations

¹ Tecnologico de Monterrey, Reserva Territorial Atlixcáyotl, Vía Atlixcáyotl No. 2301, Puebla, 72453 Mexico.
² Tecnologico de Monterrey, Carretera al Lago de Guadalupe Km. 3.5, Atizapán, Estado de Mexico 52926 México.
³ Department of Information Systems and Cyber Security, University of Texas at San Antonio, San Antonio, TX 78249 USA.

Abstract

Supervised classification based on Contrast Patterns (CP) is a trending topic in the pattern recognition literature, partly because it contains an important family of both understandable and accurate classifiers. In this paper, we survey 105 articles and provide an in-depth review of CP-based supervised classification and its applications. Based on our review, we present a taxonomy of the existing application domains of CP-based supervised classification, and a scientometric study. We also discuss potential future research opportunities.

Keywords: Contrast patterns; Review; Supervised classification; Taxonomy.