Adaptive ensemble models of extreme learning machines for time series prediction

Mark van Heeswijk; Yoan Miche; Tiina Lindh-Knuutila; Peter A.J. Hilbers; Timo Honkela; Erkki Oja; Amaury Lendasse

doi:10.1007/978-3-642-04277-5_31

Adaptive ensemble models of extreme learning machines for time series prediction

Mark van Heeswijk
, Yoan Miche
, Tiina Lindh-Knuutila
, Peter A.J. Hilbers
, Timo Honkela
, Erkki Oja
, Amaury Lendasse

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

93 Citations (Scopus)

Abstract

In this paper, we investigate the application of adaptive ensemble models of Extreme Learning Machines (ELMs) to the problem of one-step ahead prediction in (non)stationary time series. We verify that the method works on stationary time series and test the adaptivity of the ensemble model on a nonstationary time series. In the experiments, we show that the adaptive ensemble model achieves a test error comparable to the best methods, while keeping adaptivity. Moreover, it has low computational cost.

Original language	English
Title of host publication	Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2
Pages	305-314
Number of pages	10
DOIs	https://doi.org/10.1007/978-3-642-04277-5_31
Publication status	Published - 27 Nov 2009
Event	19th International Conference on Artificial Neural Networks, ICANN 2009 - Limassol, Cyprus Duration: 14 Sept 2009 → 17 Sept 2009

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	5769 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	19th International Conference on Artificial Neural Networks, ICANN 2009
Country/Territory	Cyprus
City	Limassol
Period	14/09/09 → 17/09/09

Keywords

Adaptivity
Ensemble models
Extreme learning machine
Nonstationarity
Sliding window
Time series prediction

Access to Document

10.1007/978-3-642-04277-5_31

Cite this

van Heeswijk, M., Miche, Y., Lindh-Knuutila, T., Hilbers, P. A. J., Honkela, T., Oja, E., & Lendasse, A. (2009). Adaptive ensemble models of extreme learning machines for time series prediction. In Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2 (pp. 305-314). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5769 LNCS). https://doi.org/10.1007/978-3-642-04277-5_31

van Heeswijk, Mark ; Miche, Yoan ; Lindh-Knuutila, Tiina et al. / Adaptive ensemble models of extreme learning machines for time series prediction. Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2. 2009. pp. 305-314 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{6696e81436224ab39498e3b81912a2e3,

title = "Adaptive ensemble models of extreme learning machines for time series prediction",

abstract = "In this paper, we investigate the application of adaptive ensemble models of Extreme Learning Machines (ELMs) to the problem of one-step ahead prediction in (non)stationary time series. We verify that the method works on stationary time series and test the adaptivity of the ensemble model on a nonstationary time series. In the experiments, we show that the adaptive ensemble model achieves a test error comparable to the best methods, while keeping adaptivity. Moreover, it has low computational cost.",

keywords = "Adaptivity, Ensemble models, Extreme learning machine, Nonstationarity, Sliding window, Time series prediction",

author = "\{van Heeswijk\}, Mark and Yoan Miche and Tiina Lindh-Knuutila and Hilbers, \{Peter A.J.\} and Timo Honkela and Erkki Oja and Amaury Lendasse",

year = "2009",

month = nov,

day = "27",

doi = "10.1007/978-3-642-04277-5\_31",

language = "English",

isbn = "3642042767",

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

pages = "305--314",

booktitle = "Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2",

note = "19th International Conference on Artificial Neural Networks, ICANN 2009 ; Conference date: 14-09-2009 Through 17-09-2009",

}

van Heeswijk, M, Miche, Y, Lindh-Knuutila, T, Hilbers, PAJ, Honkela, T, Oja, E & Lendasse, A 2009, Adaptive ensemble models of extreme learning machines for time series prediction. in Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5769 LNCS, pp. 305-314, 19th International Conference on Artificial Neural Networks, ICANN 2009, Limassol, Cyprus, 14/09/09. https://doi.org/10.1007/978-3-642-04277-5_31

Adaptive ensemble models of extreme learning machines for time series prediction. / van Heeswijk, Mark; Miche, Yoan; Lindh-Knuutila, Tiina et al.
Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2. 2009. p. 305-314 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5769 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Adaptive ensemble models of extreme learning machines for time series prediction

AU - van Heeswijk, Mark

AU - Miche, Yoan

AU - Lindh-Knuutila, Tiina

AU - Hilbers, Peter A.J.

AU - Honkela, Timo

AU - Oja, Erkki

AU - Lendasse, Amaury

PY - 2009/11/27

Y1 - 2009/11/27

N2 - In this paper, we investigate the application of adaptive ensemble models of Extreme Learning Machines (ELMs) to the problem of one-step ahead prediction in (non)stationary time series. We verify that the method works on stationary time series and test the adaptivity of the ensemble model on a nonstationary time series. In the experiments, we show that the adaptive ensemble model achieves a test error comparable to the best methods, while keeping adaptivity. Moreover, it has low computational cost.

AB - In this paper, we investigate the application of adaptive ensemble models of Extreme Learning Machines (ELMs) to the problem of one-step ahead prediction in (non)stationary time series. We verify that the method works on stationary time series and test the adaptivity of the ensemble model on a nonstationary time series. In the experiments, we show that the adaptive ensemble model achieves a test error comparable to the best methods, while keeping adaptivity. Moreover, it has low computational cost.

KW - Adaptivity

KW - Ensemble models

KW - Extreme learning machine

KW - Nonstationarity

KW - Sliding window

KW - Time series prediction

UR - https://www.scopus.com/pages/publications/70450194207

U2 - 10.1007/978-3-642-04277-5_31

DO - 10.1007/978-3-642-04277-5_31

M3 - Conference contribution

AN - SCOPUS:70450194207

SN - 3642042767

SN - 9783642042768

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

SP - 305

EP - 314

BT - Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2

T2 - 19th International Conference on Artificial Neural Networks, ICANN 2009

Y2 - 14 September 2009 through 17 September 2009

ER -

van Heeswijk M, Miche Y, Lindh-Knuutila T, Hilbers PAJ, Honkela T, Oja E et al. Adaptive ensemble models of extreme learning machines for time series prediction. In Artificial Neural Networks - ICANN 2009 - 19th International Conference, Proceedings: PART 2. 2009. p. 305-314. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-04277-5_31

Adaptive ensemble models of extreme learning machines for time series prediction

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this