SVM-based time series prediction with nonlinear dynamics methods

Francesco Camastra; Maurizio Filippone

doi:10.1007/978-3-540-74829-8_37

SVM-based time series prediction with nonlinear dynamics methods

Francesco Camastra^*, Maurizio Filippone

^*Corresponding author for this work

Statistics

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

6 Scopus citations

Abstract

A key problem in time series prediction using autoregressive models is to fix the model order, namely the number of past samples required to model the time series adequately. The estimation of the model order using cross-validation is a long process. In this paper we explore faster alternative to cross-validation, based on nonlinear dynamics methods, namely Grassberger-Procaccia, Kégl and False Nearest Neighbors algorithms. Once the model order is obtained, it is used to carry out the prediction, performed by a SVM. Experiments on three real data time series show that nonlinear dynamics methods have performances very close to the cross-validation ones.

Original language	English (US)
Title of host publication	Knowledge-Based Intelligent Information and Engineering Systems
Subtitle of host publication	KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings
Publisher	Springer Verlag
Pages	300-307
Number of pages	8
Edition	PART 3
ISBN (Print)	9783540748281
DOIs	https://doi.org/10.1007/978-3-540-74829-8_37
State	Published - 2007
Event	11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007 - Vietri sul Mare, Italy Duration: Sep 12 2007 → Sep 14 2007

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 3
Volume	4694 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007
Country/Territory	Italy
City	Vietri sul Mare
Period	09/12/07 → 09/14/07

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-540-74829-8_37

Cite this

Camastra, F., & Filippone, M. (2007). SVM-based time series prediction with nonlinear dynamics methods. In Knowledge-Based Intelligent Information and Engineering Systems: KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings (PART 3 ed., pp. 300-307). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4694 LNAI, No. PART 3). Springer Verlag. https://doi.org/10.1007/978-3-540-74829-8_37

Camastra, Francesco ; Filippone, Maurizio. / SVM-based time series prediction with nonlinear dynamics methods. Knowledge-Based Intelligent Information and Engineering Systems: KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings. PART 3. ed. Springer Verlag, 2007. pp. 300-307 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).

@inproceedings{ae3d1d0bfccb4052814a3d232d3c54dc,

title = "SVM-based time series prediction with nonlinear dynamics methods",

abstract = "A key problem in time series prediction using autoregressive models is to fix the model order, namely the number of past samples required to model the time series adequately. The estimation of the model order using cross-validation is a long process. In this paper we explore faster alternative to cross-validation, based on nonlinear dynamics methods, namely Grassberger-Procaccia, K{\'e}gl and False Nearest Neighbors algorithms. Once the model order is obtained, it is used to carry out the prediction, performed by a SVM. Experiments on three real data time series show that nonlinear dynamics methods have performances very close to the cross-validation ones.",

author = "Francesco Camastra and Maurizio Filippone",

year = "2007",

doi = "10.1007/978-3-540-74829-8_37",

language = "English (US)",

isbn = "9783540748281",

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

publisher = "Springer Verlag",

number = "PART 3",

pages = "300--307",

booktitle = "Knowledge-Based Intelligent Information and Engineering Systems",

address = "Germany",

edition = "PART 3",

note = "11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007 ; Conference date: 12-09-2007 Through 14-09-2007",

}

Camastra, F & Filippone, M 2007, SVM-based time series prediction with nonlinear dynamics methods. in Knowledge-Based Intelligent Information and Engineering Systems: KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings. PART 3 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 4694 LNAI, Springer Verlag, pp. 300-307, 11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007, Vietri sul Mare, Italy, 09/12/07. https://doi.org/10.1007/978-3-540-74829-8_37

SVM-based time series prediction with nonlinear dynamics methods. / Camastra, Francesco; Filippone, Maurizio.
Knowledge-Based Intelligent Information and Engineering Systems: KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings. PART 3. ed. Springer Verlag, 2007. p. 300-307 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4694 LNAI, No. PART 3).

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

TY - GEN

T1 - SVM-based time series prediction with nonlinear dynamics methods

AU - Camastra, Francesco

AU - Filippone, Maurizio

PY - 2007

Y1 - 2007

N2 - A key problem in time series prediction using autoregressive models is to fix the model order, namely the number of past samples required to model the time series adequately. The estimation of the model order using cross-validation is a long process. In this paper we explore faster alternative to cross-validation, based on nonlinear dynamics methods, namely Grassberger-Procaccia, Kégl and False Nearest Neighbors algorithms. Once the model order is obtained, it is used to carry out the prediction, performed by a SVM. Experiments on three real data time series show that nonlinear dynamics methods have performances very close to the cross-validation ones.

AB - A key problem in time series prediction using autoregressive models is to fix the model order, namely the number of past samples required to model the time series adequately. The estimation of the model order using cross-validation is a long process. In this paper we explore faster alternative to cross-validation, based on nonlinear dynamics methods, namely Grassberger-Procaccia, Kégl and False Nearest Neighbors algorithms. Once the model order is obtained, it is used to carry out the prediction, performed by a SVM. Experiments on three real data time series show that nonlinear dynamics methods have performances very close to the cross-validation ones.

UR - http://www.scopus.com/inward/record.url?scp=38049187930&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-74829-8_37

DO - 10.1007/978-3-540-74829-8_37

M3 - Conference contribution

AN - SCOPUS:38049187930

SN - 9783540748281

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

SP - 300

EP - 307

BT - Knowledge-Based Intelligent Information and Engineering Systems

PB - Springer Verlag

T2 - 11th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2007, and 17th Italian Workshop on Neural Networks, WIRN 2007

Y2 - 12 September 2007 through 14 September 2007

ER -

Camastra F, Filippone M. SVM-based time series prediction with nonlinear dynamics methods. In Knowledge-Based Intelligent Information and Engineering Systems: KES 2007 - WIRN 2007 - 11th International Conference, KES 2007, XVII Italian Workshop on Neural Networks, Proceedings. PART 3 ed. Springer Verlag. 2007. p. 300-307. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). doi: 10.1007/978-3-540-74829-8_37

SVM-based time series prediction with nonlinear dynamics methods

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this