Three step bacterial memetic algorithm

L. Gál, L. Kóczy, R. Lovassy

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

2 Citations (Scopus)

Abstract

In order to study the function approximation performance of Fuzzy Neural Networks built up from fuzzy J-K flip-flop neurons a new learning algorithm, the Three Step Bacterial Memetic Algorithm is proposed. Hybrid evolutionary methods that combine genetic type algorithms with "classic" local search have been applied to perform efficient global search. This novel version of the Bacterial Memetic Algorithm with Modified Operator Execution Order (BMAM) is a recently developed technique of hybrid type. This particular merger of evolutionary and gradient based algorithms combining both global and local search consists of bacterial mutation and, as a second step, the Levenberg-Marquardt (LM) method applied for each clone. This LM step saves in this way some potential solutions that could be lost otherwise after each mutation step. As a third step the LM algorithm is recalled for a few iterations for each individual of the population towards reaching the local optimum. In our novel algorithm various kinds of fast algorithm with less complexity, like Quasi-Newton algorithm, Conjugate Gradient algorithm, and two Backpropagation training algorithms: Gradient Descent and Gradient Descent with Adaptive Learning Rate and Momentum are nested in the bacterial mutation.

Original languageEnglish
Title of host publicationINES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings
Pages31-36
Number of pages6
DOIs
Publication statusPublished - 2010
Event14th International Conference on Intelligent Engineering Systems, INES 2010 - Las Palmas of Gran Canaria, Spain
Duration: May 5 2010May 7 2010

Other

Other14th International Conference on Intelligent Engineering Systems, INES 2010
CountrySpain
CityLas Palmas of Gran Canaria
Period5/5/105/7/10

Fingerprint

Flip flop circuits
Fuzzy neural networks
Backpropagation
Learning algorithms
Neurons
Mathematical operators
Momentum

ASJC Scopus subject areas

  • Artificial Intelligence
  • Information Systems

Cite this

Gál, L., Kóczy, L., & Lovassy, R. (2010). Three step bacterial memetic algorithm. In INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings (pp. 31-36). [5483817] https://doi.org/10.1109/INES.2010.5483817

Three step bacterial memetic algorithm. / Gál, L.; Kóczy, L.; Lovassy, R.

INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings. 2010. p. 31-36 5483817.

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

Gál, L, Kóczy, L & Lovassy, R 2010, Three step bacterial memetic algorithm. in INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings., 5483817, pp. 31-36, 14th International Conference on Intelligent Engineering Systems, INES 2010, Las Palmas of Gran Canaria, Spain, 5/5/10. https://doi.org/10.1109/INES.2010.5483817
Gál L, Kóczy L, Lovassy R. Three step bacterial memetic algorithm. In INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings. 2010. p. 31-36. 5483817 https://doi.org/10.1109/INES.2010.5483817
Gál, L. ; Kóczy, L. ; Lovassy, R. / Three step bacterial memetic algorithm. INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings. 2010. pp. 31-36
@inproceedings{97cb5f85ad4646eda727654a1177fd0a,
title = "Three step bacterial memetic algorithm",
abstract = "In order to study the function approximation performance of Fuzzy Neural Networks built up from fuzzy J-K flip-flop neurons a new learning algorithm, the Three Step Bacterial Memetic Algorithm is proposed. Hybrid evolutionary methods that combine genetic type algorithms with {"}classic{"} local search have been applied to perform efficient global search. This novel version of the Bacterial Memetic Algorithm with Modified Operator Execution Order (BMAM) is a recently developed technique of hybrid type. This particular merger of evolutionary and gradient based algorithms combining both global and local search consists of bacterial mutation and, as a second step, the Levenberg-Marquardt (LM) method applied for each clone. This LM step saves in this way some potential solutions that could be lost otherwise after each mutation step. As a third step the LM algorithm is recalled for a few iterations for each individual of the population towards reaching the local optimum. In our novel algorithm various kinds of fast algorithm with less complexity, like Quasi-Newton algorithm, Conjugate Gradient algorithm, and two Backpropagation training algorithms: Gradient Descent and Gradient Descent with Adaptive Learning Rate and Momentum are nested in the bacterial mutation.",
author = "L. G{\'a}l and L. K{\'o}czy and R. Lovassy",
year = "2010",
doi = "10.1109/INES.2010.5483817",
language = "English",
isbn = "9781424476527",
pages = "31--36",
booktitle = "INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings",

}

TY - GEN

T1 - Three step bacterial memetic algorithm

AU - Gál, L.

AU - Kóczy, L.

AU - Lovassy, R.

PY - 2010

Y1 - 2010

N2 - In order to study the function approximation performance of Fuzzy Neural Networks built up from fuzzy J-K flip-flop neurons a new learning algorithm, the Three Step Bacterial Memetic Algorithm is proposed. Hybrid evolutionary methods that combine genetic type algorithms with "classic" local search have been applied to perform efficient global search. This novel version of the Bacterial Memetic Algorithm with Modified Operator Execution Order (BMAM) is a recently developed technique of hybrid type. This particular merger of evolutionary and gradient based algorithms combining both global and local search consists of bacterial mutation and, as a second step, the Levenberg-Marquardt (LM) method applied for each clone. This LM step saves in this way some potential solutions that could be lost otherwise after each mutation step. As a third step the LM algorithm is recalled for a few iterations for each individual of the population towards reaching the local optimum. In our novel algorithm various kinds of fast algorithm with less complexity, like Quasi-Newton algorithm, Conjugate Gradient algorithm, and two Backpropagation training algorithms: Gradient Descent and Gradient Descent with Adaptive Learning Rate and Momentum are nested in the bacterial mutation.

AB - In order to study the function approximation performance of Fuzzy Neural Networks built up from fuzzy J-K flip-flop neurons a new learning algorithm, the Three Step Bacterial Memetic Algorithm is proposed. Hybrid evolutionary methods that combine genetic type algorithms with "classic" local search have been applied to perform efficient global search. This novel version of the Bacterial Memetic Algorithm with Modified Operator Execution Order (BMAM) is a recently developed technique of hybrid type. This particular merger of evolutionary and gradient based algorithms combining both global and local search consists of bacterial mutation and, as a second step, the Levenberg-Marquardt (LM) method applied for each clone. This LM step saves in this way some potential solutions that could be lost otherwise after each mutation step. As a third step the LM algorithm is recalled for a few iterations for each individual of the population towards reaching the local optimum. In our novel algorithm various kinds of fast algorithm with less complexity, like Quasi-Newton algorithm, Conjugate Gradient algorithm, and two Backpropagation training algorithms: Gradient Descent and Gradient Descent with Adaptive Learning Rate and Momentum are nested in the bacterial mutation.

UR - http://www.scopus.com/inward/record.url?scp=77954805251&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954805251&partnerID=8YFLogxK

U2 - 10.1109/INES.2010.5483817

DO - 10.1109/INES.2010.5483817

M3 - Conference contribution

AN - SCOPUS:77954805251

SN - 9781424476527

SP - 31

EP - 36

BT - INES 2010 - 14th International Conference on Intelligent Engineering Systems, Proceedings

ER -