A survey on concept drift adaptation

J. Gama, I. Zliobaite, A. Bifet, M. Pechenizkiy, A. Bouchachia

Research output: Contribution to journalArticleAcademicpeer-review

675 Citations (Scopus)
1 Downloads (Pure)

Abstract

Concept drift primarily refers to an online supervised learning scenario when the relation between the input data and the target variable changes over time. Assuming a general knowledge of supervised learning in this article, we characterize adaptive learning processes; categorize existing strategies for handling concept drift; overview the most representative, distinct, and popular techniques and algorithms; discuss evaluation methodology of adaptive algorithms; and present a set of illustrative applications. The survey covers the different facets of concept drift in an integrated way to reflect on the existing scattered state of the art. Thus, it aims at providing a comprehensive introduction to the concept drift adaptation for researchers, industry analysts, and practitioners.
Original languageEnglish
Article number44
Pages (from-to)44/1-37
Number of pages37
JournalACM Computing Surveys
Volume46
Issue number4
DOIs
Publication statusPublished - 2014

Fingerprint

Concept Drift
Supervised learning
Supervised Learning
Adaptive algorithms
Adaptive Processes
Adaptive Learning
Online Learning
Adaptive Algorithm
Learning Process
Facet
Industry
Cover
Distinct
Scenarios
Target
Methodology
Evaluation

Cite this

Gama, J., Zliobaite, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014). A survey on concept drift adaptation. ACM Computing Surveys, 46(4), 44/1-37. [44]. https://doi.org/10.1145/2523813
Gama, J. ; Zliobaite, I. ; Bifet, A. ; Pechenizkiy, M. ; Bouchachia, A. / A survey on concept drift adaptation. In: ACM Computing Surveys. 2014 ; Vol. 46, No. 4. pp. 44/1-37.
@article{2866e3349a5c4fad85ece44e075a6365,
title = "A survey on concept drift adaptation",
abstract = "Concept drift primarily refers to an online supervised learning scenario when the relation between the input data and the target variable changes over time. Assuming a general knowledge of supervised learning in this article, we characterize adaptive learning processes; categorize existing strategies for handling concept drift; overview the most representative, distinct, and popular techniques and algorithms; discuss evaluation methodology of adaptive algorithms; and present a set of illustrative applications. The survey covers the different facets of concept drift in an integrated way to reflect on the existing scattered state of the art. Thus, it aims at providing a comprehensive introduction to the concept drift adaptation for researchers, industry analysts, and practitioners.",
author = "J. Gama and I. Zliobaite and A. Bifet and M. Pechenizkiy and A. Bouchachia",
year = "2014",
doi = "10.1145/2523813",
language = "English",
volume = "46",
pages = "44/1--37",
journal = "ACM Computing Surveys",
issn = "0360-0300",
publisher = "Association for Computing Machinery, Inc",
number = "4",

}

Gama, J, Zliobaite, I, Bifet, A, Pechenizkiy, M & Bouchachia, A 2014, 'A survey on concept drift adaptation', ACM Computing Surveys, vol. 46, no. 4, 44, pp. 44/1-37. https://doi.org/10.1145/2523813

A survey on concept drift adaptation. / Gama, J.; Zliobaite, I.; Bifet, A.; Pechenizkiy, M.; Bouchachia, A.

In: ACM Computing Surveys, Vol. 46, No. 4, 44, 2014, p. 44/1-37.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - A survey on concept drift adaptation

AU - Gama, J.

AU - Zliobaite, I.

AU - Bifet, A.

AU - Pechenizkiy, M.

AU - Bouchachia, A.

PY - 2014

Y1 - 2014

N2 - Concept drift primarily refers to an online supervised learning scenario when the relation between the input data and the target variable changes over time. Assuming a general knowledge of supervised learning in this article, we characterize adaptive learning processes; categorize existing strategies for handling concept drift; overview the most representative, distinct, and popular techniques and algorithms; discuss evaluation methodology of adaptive algorithms; and present a set of illustrative applications. The survey covers the different facets of concept drift in an integrated way to reflect on the existing scattered state of the art. Thus, it aims at providing a comprehensive introduction to the concept drift adaptation for researchers, industry analysts, and practitioners.

AB - Concept drift primarily refers to an online supervised learning scenario when the relation between the input data and the target variable changes over time. Assuming a general knowledge of supervised learning in this article, we characterize adaptive learning processes; categorize existing strategies for handling concept drift; overview the most representative, distinct, and popular techniques and algorithms; discuss evaluation methodology of adaptive algorithms; and present a set of illustrative applications. The survey covers the different facets of concept drift in an integrated way to reflect on the existing scattered state of the art. Thus, it aims at providing a comprehensive introduction to the concept drift adaptation for researchers, industry analysts, and practitioners.

U2 - 10.1145/2523813

DO - 10.1145/2523813

M3 - Article

VL - 46

SP - 44/1-37

JO - ACM Computing Surveys

JF - ACM Computing Surveys

SN - 0360-0300

IS - 4

M1 - 44

ER -

Gama J, Zliobaite I, Bifet A, Pechenizkiy M, Bouchachia A. A survey on concept drift adaptation. ACM Computing Surveys. 2014;46(4):44/1-37. 44. https://doi.org/10.1145/2523813