Use of backpropagation and differential evolution algorithms to training MLPs

Luiz Carlos Camargo; Hegler Correa Tissot; Aurora Trinidad Ramirez Pozo

doi:10.1109/SCCC.2012.17

Use of backpropagation and differential evolution algorithms to training MLPs

Luiz Carlos Camargo, Hegler Correa Tissot, Aurora Trinidad Ramirez Pozo

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

1 Citation (Scopus)

Abstract

Artificial Neural Networks (ANNs) are often used (trained) to find a general solution in problems where a pattern needs to be extracted, such as data classification. Feedforward (FFNN) is one of the ANN architectures and multilayer perceptron (MLP) is a type of FFNN. Based on gradient descent, backpropagation (BP) is one of the most used algorithms for MLP training. Evolutionary algorithms can be also used to train MLPs, including Differential Evolution (DE) algorithm. In this paper, BP and DE are used to train MLPs and they are both compared in four different approaches: (a) backpropagation, (b) DE with fixed parameter values, (c) DE with adaptive parameter values and (d) a hybrid alternative using both DE+BP algorithms.

Original language	English
Title of host publication	Proceedings - 31st International Conference of the Chilean Computer Science Society, SCCC 2012
Place of Publication	Pistcataway, NJ
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	78-86
Number of pages	9
ISBN (Print)	9781479929375
DOIs	https://doi.org/10.1109/SCCC.2012.17
Publication status	Published - 2013
Externally published	Yes
Event	31st International Conference of the Chilean Computer Science Society, SCCC 2012 - Valparaiso, Chile Duration: 12 Nov 2012 → 16 Nov 2012

Other

Other	31st International Conference of the Chilean Computer Science Society, SCCC 2012
Country/Territory	Chile
City	Valparaiso
Period	12/11/12 → 16/11/12

Keywords

Artificial Neural Network
Backpropagation (BP) algorithm
Differential evolution (DE) algorithm
Multilayer perceptron

Access to Document

10.1109/SCCC.2012.17

Cite this

@inproceedings{7b612d00a68c4afd902b74578c0751ee,

title = "Use of backpropagation and differential evolution algorithms to training MLPs",

abstract = "Artificial Neural Networks (ANNs) are often used (trained) to find a general solution in problems where a pattern needs to be extracted, such as data classification. Feedforward (FFNN) is one of the ANN architectures and multilayer perceptron (MLP) is a type of FFNN. Based on gradient descent, backpropagation (BP) is one of the most used algorithms for MLP training. Evolutionary algorithms can be also used to train MLPs, including Differential Evolution (DE) algorithm. In this paper, BP and DE are used to train MLPs and they are both compared in four different approaches: (a) backpropagation, (b) DE with fixed parameter values, (c) DE with adaptive parameter values and (d) a hybrid alternative using both DE+BP algorithms.",

keywords = "Artificial Neural Network, Backpropagation (BP) algorithm, Differential evolution (DE) algorithm, Multilayer perceptron",

author = "Camargo, {Luiz Carlos} and Tissot, {Hegler Correa} and Pozo, {Aurora Trinidad Ramirez}",

year = "2013",

doi = "10.1109/SCCC.2012.17",

language = "English",

isbn = "9781479929375",

pages = "78--86",

booktitle = "Proceedings - 31st International Conference of the Chilean Computer Science Society, SCCC 2012",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

note = "31st International Conference of the Chilean Computer Science Society, SCCC 2012 ; Conference date: 12-11-2012 Through 16-11-2012",

}

Camargo, LC, Tissot, HC & Pozo, ATR 2013, Use of backpropagation and differential evolution algorithms to training MLPs. in Proceedings - 31st International Conference of the Chilean Computer Science Society, SCCC 2012. Institute of Electrical and Electronics Engineers (IEEE), Pistcataway, NJ, pp. 78-86, 31st International Conference of the Chilean Computer Science Society, SCCC 2012, Valparaiso, Chile, 12/11/12. https://doi.org/10.1109/SCCC.2012.17

Use of backpropagation and differential evolution algorithms to training MLPs. / Camargo, Luiz Carlos; Tissot, Hegler Correa; Pozo, Aurora Trinidad Ramirez.
Proceedings - 31st International Conference of the Chilean Computer Science Society, SCCC 2012. Pistcataway, NJ: Institute of Electrical and Electronics Engineers (IEEE), 2013. p. 78-86.

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

TY - GEN

T1 - Use of backpropagation and differential evolution algorithms to training MLPs

AU - Camargo, Luiz Carlos

AU - Tissot, Hegler Correa

AU - Pozo, Aurora Trinidad Ramirez

PY - 2013

Y1 - 2013

N2 - Artificial Neural Networks (ANNs) are often used (trained) to find a general solution in problems where a pattern needs to be extracted, such as data classification. Feedforward (FFNN) is one of the ANN architectures and multilayer perceptron (MLP) is a type of FFNN. Based on gradient descent, backpropagation (BP) is one of the most used algorithms for MLP training. Evolutionary algorithms can be also used to train MLPs, including Differential Evolution (DE) algorithm. In this paper, BP and DE are used to train MLPs and they are both compared in four different approaches: (a) backpropagation, (b) DE with fixed parameter values, (c) DE with adaptive parameter values and (d) a hybrid alternative using both DE+BP algorithms.

AB - Artificial Neural Networks (ANNs) are often used (trained) to find a general solution in problems where a pattern needs to be extracted, such as data classification. Feedforward (FFNN) is one of the ANN architectures and multilayer perceptron (MLP) is a type of FFNN. Based on gradient descent, backpropagation (BP) is one of the most used algorithms for MLP training. Evolutionary algorithms can be also used to train MLPs, including Differential Evolution (DE) algorithm. In this paper, BP and DE are used to train MLPs and they are both compared in four different approaches: (a) backpropagation, (b) DE with fixed parameter values, (c) DE with adaptive parameter values and (d) a hybrid alternative using both DE+BP algorithms.

KW - Artificial Neural Network

KW - Backpropagation (BP) algorithm

KW - Differential evolution (DE) algorithm

KW - Multilayer perceptron

UR - http://www.scopus.com/inward/record.url?scp=84893579248&partnerID=8YFLogxK

U2 - 10.1109/SCCC.2012.17

DO - 10.1109/SCCC.2012.17

M3 - Conference proceeding contribution

SN - 9781479929375

SP - 78

EP - 86

BT - Proceedings - 31st International Conference of the Chilean Computer Science Society, SCCC 2012

PB - Institute of Electrical and Electronics Engineers (IEEE)

CY - Pistcataway, NJ

T2 - 31st International Conference of the Chilean Computer Science Society, SCCC 2012

Y2 - 12 November 2012 through 16 November 2012

ER -

Use of backpropagation and differential evolution algorithms to training MLPs

Abstract

Other

Keywords

Access to Document

Other files and links

Fingerprint

Cite this