Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations

Tamás Dózsa; Gergö Bognar; Peter Kovacs

doi:10.1007/978-3-030-45096-0_44

Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations

Tamás Dózsa
, Gergö Bognar
, Peter Kovacs

Institut für Signalverarbeitung

Publikation: Beitrag in Buch/Bericht/Konferenzband › Konferenzbeitrag › Begutachtung

Abstract

In this work, we are focusing on the problem of heartbeat classification in electrocardiogram (ECG) signals. First we develop a patient-specific feature extraction scheme by using adaptive orthogonal transformations based on wavelets, B-splines, Hermite and rational functions. The so-called variable projection provides the general framework to find the optimal nonlinear parameters of these transformations. After extracting the features, we train a support vector machine (SVM) for each model whose outputs are combined via ensemble learning techniques. In the experiments, we achieved an accuracy of 94.2% on the PhysioNet MIT-BIH Arrhythmia Database that shows the potential of the proposed signal models in arrhythmia detection.

Originalsprache	Englisch
Titel	Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers
Herausgeber*innen	Roberto Moreno-Díaz, Alexis Quesada-Arencibia, Franz Pichler
Verlag	Springer
Seiten	355-363
Seitenumfang	9
ISBN (elektronisch)	978-3-030-45096-0
ISBN (Print)	9783030450953
DOIs	https://doi.org/10.1007/978-3-030-45096-0_44
Publikationsstatus	Veröffentlicht - Apr. 2020

Publikationsreihe

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band	12014 LNCS
ISSN (Print)	0302-9743
ISSN (elektronisch)	1611-3349

Wissenschaftszweige

102019 Machine Learning
202 Elektrotechnik, Elektronik, Informationstechnik
202022 Informationstechnik
202037 Signalverarbeitung

JKU-Schwerpunkte

Digital Transformation

Zugriff auf Dokument

10.1007/978-3-030-45096-0_44

Andere Dateien und Links

Morphology based Cardiac Signal Processing
Böck, C. (Forscher*in), Lunglmayr, M. (Forscher*in) & Huemer, M. (Projektleiter*in)
01.04.2016 → 31.03.2020
Projekt: Anderes › Projekt aus Wissenschaftsgebiet der Forschungseinheit

Dieses zitieren

Dózsa, T., Bognar, G., & Kovacs, P. (2020). Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations. in R. Moreno-Díaz, A. Quesada-Arencibia, & F. Pichler (Hrsg.), Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers (S. 355-363). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 12014 LNCS). Springer. https://doi.org/10.1007/978-3-030-45096-0_44

Dózsa, Tamás ; Bognar, Gergö ; Kovacs, Peter. / Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations. Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers. Hrsg. / Roberto Moreno-Díaz ; Alexis Quesada-Arencibia ; Franz Pichler. Springer, 2020. S. 355-363 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{359c89ee1ea241fcb3003e36ef786a3a,

title = "Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations",

abstract = "In this work, we are focusing on the problem of heartbeat classification in electrocardiogram (ECG) signals. First we develop a patient-specific feature extraction scheme by using adaptive orthogonal transformations based on wavelets, B-splines, Hermite and rational functions. The so-called variable projection provides the general framework to find the optimal nonlinear parameters of these transformations. After extracting the features, we train a support vector machine (SVM) for each model whose outputs are combined via ensemble learning techniques. In the experiments, we achieved an accuracy of 94.2\% on the PhysioNet MIT-BIH Arrhythmia Database that shows the potential of the proposed signal models in arrhythmia detection.",

author = "Tam{\'a}s D{\'o}zsa and Gerg{\"o} Bognar and Peter Kovacs",

year = "2020",

month = apr,

doi = "10.1007/978-3-030-45096-0\_44",

language = "English",

isbn = "9783030450953",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "355--363",

editor = "Roberto Moreno-D{\'i}az and Alexis Quesada-Arencibia and Franz Pichler",

booktitle = "Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers",

}

Dózsa, T, Bognar, G & Kovacs, P 2020, Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations. in R Moreno-Díaz, A Quesada-Arencibia & F Pichler (Hrsg.), Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bd. 12014 LNCS, Springer, S. 355-363. https://doi.org/10.1007/978-3-030-45096-0_44

Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations. / Dózsa, Tamás; Bognar, Gergö; Kovacs, Peter.
Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers. Hrsg. / Roberto Moreno-Díaz; Alexis Quesada-Arencibia; Franz Pichler. Springer, 2020. S. 355-363 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 12014 LNCS).

Publikation: Beitrag in Buch/Bericht/Konferenzband › Konferenzbeitrag › Begutachtung

TY - GEN

T1 - Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations

AU - Dózsa, Tamás

AU - Bognar, Gergö

AU - Kovacs, Peter

PY - 2020/4

Y1 - 2020/4

N2 - In this work, we are focusing on the problem of heartbeat classification in electrocardiogram (ECG) signals. First we develop a patient-specific feature extraction scheme by using adaptive orthogonal transformations based on wavelets, B-splines, Hermite and rational functions. The so-called variable projection provides the general framework to find the optimal nonlinear parameters of these transformations. After extracting the features, we train a support vector machine (SVM) for each model whose outputs are combined via ensemble learning techniques. In the experiments, we achieved an accuracy of 94.2% on the PhysioNet MIT-BIH Arrhythmia Database that shows the potential of the proposed signal models in arrhythmia detection.

AB - In this work, we are focusing on the problem of heartbeat classification in electrocardiogram (ECG) signals. First we develop a patient-specific feature extraction scheme by using adaptive orthogonal transformations based on wavelets, B-splines, Hermite and rational functions. The so-called variable projection provides the general framework to find the optimal nonlinear parameters of these transformations. After extracting the features, we train a support vector machine (SVM) for each model whose outputs are combined via ensemble learning techniques. In the experiments, we achieved an accuracy of 94.2% on the PhysioNet MIT-BIH Arrhythmia Database that shows the potential of the proposed signal models in arrhythmia detection.

UR - https://www.springer.com/gp/book/9783030450953

UR - https://www.scopus.com/pages/publications/85083991096

U2 - 10.1007/978-3-030-45096-0_44

DO - 10.1007/978-3-030-45096-0_44

M3 - Conference proceedings

SN - 9783030450953

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 355

EP - 363

BT - Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers

A2 - Moreno-Díaz, Roberto

A2 - Quesada-Arencibia, Alexis

A2 - Pichler, Franz

PB - Springer

ER -

Dózsa T, Bognar G, Kovacs P. Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations. in Moreno-Díaz R, Quesada-Arencibia A, Pichler F, Hrsg., Computer Aided Systems Theory – EUROCAST 2019 - 17th International Conference, Revised Selected Papers. Springer. 2020. S. 355-363. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-45096-0_44

Ensemble Learning for Heartbeat Classification Using Adaptive Orthogonal Transformations

Abstract

Publikationsreihe

Wissenschaftszweige

JKU-Schwerpunkte

Zugriff auf Dokument

Andere Dateien und Links

Projekte

Morphology based Cardiac Signal Processing

Dieses zitieren