A cluster based prototype reduction for online classification

Kemilly Dearo Garcia, André C.P.L.F. de Carvalho, Joao Mendes-Moreira

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)

Abstract

Data stream is a challenging research topic in which data can continuously arrive with a probability distribution that may change over time. Depending on the changes in the data distribution, different phenomena can occur, for example, a concept drift. A concept drift occurs when the concepts associated with a dataset change when new data arrive. This paper proposes a new method based on k-Nearest Neighbors that implements a sliding window requiring less instances stored for training than existing methods. For such, a clustering approach is used to summarize data by placing labeled instances considered similar in the same cluster. Besides, instances close to the uncertainty border of existing classes are also stored, in a sliding window, to adapt the model to concept drift. The proposed method is experimentally compared with state-of-the-art classifiers from the data stream literature, regarding accuracy and processing time. According to the experimental results, the proposed method has better accuracy and less time consumption when fewer information about the concepts are stored in a single sliding window.
Original languageEnglish
Title of host publicationIntelligent Data Engineering and Automated Learning - IDEAL 2018
Subtitle of host publication19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I
EditorsHujun Yin, David Camacho, Paulo Novais, Antonio J. Tallón-Ballesteros
PublisherSpringer
Pages603-610
Number of pages8
ISBN (Electronic)978-3-030-03493-1
ISBN (Print)978-3-030-03492-4
DOIs
Publication statusE-pub ahead of print/First online - 9 Nov 2018
Event19th International Conference on Intelligent Data Engineering and Automated Learning 2018 - Autonomous University of Madrid, Madrid, Spain
Duration: 21 Nov 201823 Nov 2018
Conference number: 19
https://aida.ii.uam.es/ideal2018/#!/main

Publication series

NameLecture Notes in Computer Science
PublisherSpringer, Cham
Volume11314
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference19th International Conference on Intelligent Data Engineering and Automated Learning 2018
Abbreviated titleIDEAL 2018
CountrySpain
CityMadrid
Period21/11/1823/11/18
Internet address

Fingerprint

Probability distributions
Classifiers
Processing
Uncertainty

Cite this

Dearo Garcia, K., de Carvalho, A. C. P. L. F., & Mendes-Moreira, J. (2018). A cluster based prototype reduction for online classification. In H. Yin, D. Camacho, P. Novais, & A. J. Tallón-Ballesteros (Eds.), Intelligent Data Engineering and Automated Learning - IDEAL 2018: 19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I (pp. 603-610). (Lecture Notes in Computer Science; Vol. 11314). Springer. https://doi.org/10.1007/978-3-030-03493-1
Dearo Garcia, Kemilly ; de Carvalho, André C.P.L.F. ; Mendes-Moreira, Joao. / A cluster based prototype reduction for online classification. Intelligent Data Engineering and Automated Learning - IDEAL 2018: 19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I. editor / Hujun Yin ; David Camacho ; Paulo Novais ; Antonio J. Tallón-Ballesteros. Springer, 2018. pp. 603-610 (Lecture Notes in Computer Science).
@inproceedings{e204efb204a74746be8aeb3e34883bd3,
title = "A cluster based prototype reduction for online classification",
abstract = "Data stream is a challenging research topic in which data can continuously arrive with a probability distribution that may change over time. Depending on the changes in the data distribution, different phenomena can occur, for example, a concept drift. A concept drift occurs when the concepts associated with a dataset change when new data arrive. This paper proposes a new method based on k-Nearest Neighbors that implements a sliding window requiring less instances stored for training than existing methods. For such, a clustering approach is used to summarize data by placing labeled instances considered similar in the same cluster. Besides, instances close to the uncertainty border of existing classes are also stored, in a sliding window, to adapt the model to concept drift. The proposed method is experimentally compared with state-of-the-art classifiers from the data stream literature, regarding accuracy and processing time. According to the experimental results, the proposed method has better accuracy and less time consumption when fewer information about the concepts are stored in a single sliding window.",
author = "{Dearo Garcia}, Kemilly and {de Carvalho}, {Andr{\'e} C.P.L.F.} and Joao Mendes-Moreira",
year = "2018",
month = "11",
day = "9",
doi = "10.1007/978-3-030-03493-1",
language = "English",
isbn = "978-3-030-03492-4",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "603--610",
editor = "Hujun Yin and David Camacho and Paulo Novais and Tall{\'o}n-Ballesteros, {Antonio J.}",
booktitle = "Intelligent Data Engineering and Automated Learning - IDEAL 2018",

}

Dearo Garcia, K, de Carvalho, ACPLF & Mendes-Moreira, J 2018, A cluster based prototype reduction for online classification. in H Yin, D Camacho, P Novais & AJ Tallón-Ballesteros (eds), Intelligent Data Engineering and Automated Learning - IDEAL 2018: 19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I. Lecture Notes in Computer Science, vol. 11314, Springer, pp. 603-610, 19th International Conference on Intelligent Data Engineering and Automated Learning 2018, Madrid, Spain, 21/11/18. https://doi.org/10.1007/978-3-030-03493-1

A cluster based prototype reduction for online classification. / Dearo Garcia, Kemilly ; de Carvalho, André C.P.L.F.; Mendes-Moreira, Joao.

Intelligent Data Engineering and Automated Learning - IDEAL 2018: 19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I. ed. / Hujun Yin; David Camacho; Paulo Novais; Antonio J. Tallón-Ballesteros. Springer, 2018. p. 603-610 (Lecture Notes in Computer Science; Vol. 11314).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - A cluster based prototype reduction for online classification

AU - Dearo Garcia, Kemilly

AU - de Carvalho, André C.P.L.F.

AU - Mendes-Moreira, Joao

PY - 2018/11/9

Y1 - 2018/11/9

N2 - Data stream is a challenging research topic in which data can continuously arrive with a probability distribution that may change over time. Depending on the changes in the data distribution, different phenomena can occur, for example, a concept drift. A concept drift occurs when the concepts associated with a dataset change when new data arrive. This paper proposes a new method based on k-Nearest Neighbors that implements a sliding window requiring less instances stored for training than existing methods. For such, a clustering approach is used to summarize data by placing labeled instances considered similar in the same cluster. Besides, instances close to the uncertainty border of existing classes are also stored, in a sliding window, to adapt the model to concept drift. The proposed method is experimentally compared with state-of-the-art classifiers from the data stream literature, regarding accuracy and processing time. According to the experimental results, the proposed method has better accuracy and less time consumption when fewer information about the concepts are stored in a single sliding window.

AB - Data stream is a challenging research topic in which data can continuously arrive with a probability distribution that may change over time. Depending on the changes in the data distribution, different phenomena can occur, for example, a concept drift. A concept drift occurs when the concepts associated with a dataset change when new data arrive. This paper proposes a new method based on k-Nearest Neighbors that implements a sliding window requiring less instances stored for training than existing methods. For such, a clustering approach is used to summarize data by placing labeled instances considered similar in the same cluster. Besides, instances close to the uncertainty border of existing classes are also stored, in a sliding window, to adapt the model to concept drift. The proposed method is experimentally compared with state-of-the-art classifiers from the data stream literature, regarding accuracy and processing time. According to the experimental results, the proposed method has better accuracy and less time consumption when fewer information about the concepts are stored in a single sliding window.

U2 - 10.1007/978-3-030-03493-1

DO - 10.1007/978-3-030-03493-1

M3 - Conference contribution

SN - 978-3-030-03492-4

T3 - Lecture Notes in Computer Science

SP - 603

EP - 610

BT - Intelligent Data Engineering and Automated Learning - IDEAL 2018

A2 - Yin, Hujun

A2 - Camacho, David

A2 - Novais, Paulo

A2 - Tallón-Ballesteros, Antonio J.

PB - Springer

ER -

Dearo Garcia K, de Carvalho ACPLF, Mendes-Moreira J. A cluster based prototype reduction for online classification. In Yin H, Camacho D, Novais P, Tallón-Ballesteros AJ, editors, Intelligent Data Engineering and Automated Learning - IDEAL 2018: 19th International Conference, Madrid, Spain, November 21–23, 2018, Proceedings, Part I. Springer. 2018. p. 603-610. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-030-03493-1