Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks

Víctor Manuel Landassuri-Moreno, Carmen L. Bustillo-Hernández, José Juan Carbajal-Hernández, Luis P. Sańchez Fernández

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

In recent years, Evolutionary Algorithms (EAs) have been remarkably useful to improve the robustness of Artificial Neural Networks (ANNs). This study introduces an experimental analysis using an EAs aimed to evolve ANNs architectures (the FS-EPNet algorithm) to understand how neural networks are evolved with a steady-state algorithm and compare the Single-step-ahead (SSP) and Multiple-step-ahead (MSP) methods for prediction tasks over two test sets. It was decided to test an inside-set during evolution and an outside-set after the whole evolutionary process has been completed to validate the generalization performance with the same method (SSP or MSP). Thus, the networks may not be correctly evaluated (misleading fitness) if the single SSP is used during evolution (inside-set) and then the MSP at the end of it (outside-set). The results show that the same prediction method should be used in both evaluation sets providing smaller errors on average. © Springer-Verlag 2013.
Original languageAmerican English
Title of host publicationSingle-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks
Pages65-72
Number of pages57
ISBN (Electronic)9783642418211
DOIs
StatePublished - 1 Dec 2013
EventLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) -
Duration: 1 Jan 2014 → …

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8258 LNCS
ISSN (Print)0302-9743

Conference

ConferenceLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Period1/01/14 → …

Fingerprint

Evolutionary Neural Networks
Artificial Neural Network
Neural networks
Evolutionary algorithms
Prediction
Evolutionary Algorithms
Network architecture
Experimental Analysis
Test Set
Network Architecture
Fitness
Neural Networks
Robustness
Evaluation

Cite this

Landassuri-Moreno, V. M., Bustillo-Hernández, C. L., Carbajal-Hernández, J. J., & Sańchez Fernández, L. P. (2013). Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. In Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks (pp. 65-72). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8258 LNCS). https://doi.org/10.1007/978-3-642-41822-8_9
Landassuri-Moreno, Víctor Manuel ; Bustillo-Hernández, Carmen L. ; Carbajal-Hernández, José Juan ; Sańchez Fernández, Luis P. / Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. 2013. pp. 65-72 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{2626e01c6d374e00a95c8b72b1426018,
title = "Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks",
abstract = "In recent years, Evolutionary Algorithms (EAs) have been remarkably useful to improve the robustness of Artificial Neural Networks (ANNs). This study introduces an experimental analysis using an EAs aimed to evolve ANNs architectures (the FS-EPNet algorithm) to understand how neural networks are evolved with a steady-state algorithm and compare the Single-step-ahead (SSP) and Multiple-step-ahead (MSP) methods for prediction tasks over two test sets. It was decided to test an inside-set during evolution and an outside-set after the whole evolutionary process has been completed to validate the generalization performance with the same method (SSP or MSP). Thus, the networks may not be correctly evaluated (misleading fitness) if the single SSP is used during evolution (inside-set) and then the MSP at the end of it (outside-set). The results show that the same prediction method should be used in both evaluation sets providing smaller errors on average. {\circledC} Springer-Verlag 2013.",
author = "Landassuri-Moreno, {V{\'i}ctor Manuel} and Bustillo-Hern{\'a}ndez, {Carmen L.} and Carbajal-Hern{\'a}ndez, {Jos{\'e} Juan} and {Sańchez Fern{\'a}ndez}, {Luis P.}",
year = "2013",
month = "12",
day = "1",
doi = "10.1007/978-3-642-41822-8_9",
language = "American English",
isbn = "9783642418211",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "65--72",
booktitle = "Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks",

}

Landassuri-Moreno, VM, Bustillo-Hernández, CL, Carbajal-Hernández, JJ & Sańchez Fernández, LP 2013, Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. in Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8258 LNCS, pp. 65-72, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 1/01/14. https://doi.org/10.1007/978-3-642-41822-8_9

Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. / Landassuri-Moreno, Víctor Manuel; Bustillo-Hernández, Carmen L.; Carbajal-Hernández, José Juan; Sańchez Fernández, Luis P.

Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. 2013. p. 65-72 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8258 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks

AU - Landassuri-Moreno, Víctor Manuel

AU - Bustillo-Hernández, Carmen L.

AU - Carbajal-Hernández, José Juan

AU - Sańchez Fernández, Luis P.

PY - 2013/12/1

Y1 - 2013/12/1

N2 - In recent years, Evolutionary Algorithms (EAs) have been remarkably useful to improve the robustness of Artificial Neural Networks (ANNs). This study introduces an experimental analysis using an EAs aimed to evolve ANNs architectures (the FS-EPNet algorithm) to understand how neural networks are evolved with a steady-state algorithm and compare the Single-step-ahead (SSP) and Multiple-step-ahead (MSP) methods for prediction tasks over two test sets. It was decided to test an inside-set during evolution and an outside-set after the whole evolutionary process has been completed to validate the generalization performance with the same method (SSP or MSP). Thus, the networks may not be correctly evaluated (misleading fitness) if the single SSP is used during evolution (inside-set) and then the MSP at the end of it (outside-set). The results show that the same prediction method should be used in both evaluation sets providing smaller errors on average. © Springer-Verlag 2013.

AB - In recent years, Evolutionary Algorithms (EAs) have been remarkably useful to improve the robustness of Artificial Neural Networks (ANNs). This study introduces an experimental analysis using an EAs aimed to evolve ANNs architectures (the FS-EPNet algorithm) to understand how neural networks are evolved with a steady-state algorithm and compare the Single-step-ahead (SSP) and Multiple-step-ahead (MSP) methods for prediction tasks over two test sets. It was decided to test an inside-set during evolution and an outside-set after the whole evolutionary process has been completed to validate the generalization performance with the same method (SSP or MSP). Thus, the networks may not be correctly evaluated (misleading fitness) if the single SSP is used during evolution (inside-set) and then the MSP at the end of it (outside-set). The results show that the same prediction method should be used in both evaluation sets providing smaller errors on average. © Springer-Verlag 2013.

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84893190926&origin=inward

UR - https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=84893190926&origin=inward

U2 - 10.1007/978-3-642-41822-8_9

DO - 10.1007/978-3-642-41822-8_9

M3 - Conference contribution

SN - 9783642418211

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

SP - 65

EP - 72

BT - Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks

ER -

Landassuri-Moreno VM, Bustillo-Hernández CL, Carbajal-Hernández JJ, Sańchez Fernández LP. Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. In Single-step-ahead and multi-step-ahead prediction with evolutionary artificial neural networks. 2013. p. 65-72. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-41822-8_9