Using genetic algorithm with adaptive mutation mechanism for neural networks design and training

Y. R. Tsoy, V. G. Spitsyn

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

9 Citations (Scopus)

Abstract

In this paper a developed evolutionary algorithm (NEvA) for simultaneous connections and weights of neural network training is described. A distinctive feature of the algorithm is flexible and effective evolutionary search and a balanced resulting neural network structure due to adaptive mutation operator. In NEvA neural network structure changes, caused by mutation operator, as well as mutation rate are defined independently for each individual. Two different problems are chosen to test the algorithm. The first one is a simple 2-bit parity problem, well known as XOR problem, and the second is a neurocontrol problem of t and 2 poles balancing. A comparison of obtained results with results of other algorithms is presented.

Original languageEnglish
Title of host publicationProceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005
Pages709-714
Number of pages6
Volume1
DOIs
Publication statusPublished - 2005
Event9th Russian-Korean International Symposium on Science and Technology, KORUS-2005 - Novosibirsk, Russian Federation
Duration: 26 Jun 20052 Jul 2005

Other

Other9th Russian-Korean International Symposium on Science and Technology, KORUS-2005
CountryRussian Federation
CityNovosibirsk
Period26.6.052.7.05

Fingerprint

Genetic algorithms
Neural networks
Evolutionary algorithms
Poles

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Tsoy, Y. R., & Spitsyn, V. G. (2005). Using genetic algorithm with adaptive mutation mechanism for neural networks design and training. In Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005 (Vol. 1, pp. 709-714). [1507882] https://doi.org/10.1109/KORUS.2005.1507882

Using genetic algorithm with adaptive mutation mechanism for neural networks design and training. / Tsoy, Y. R.; Spitsyn, V. G.

Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005. Vol. 1 2005. p. 709-714 1507882.

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

Tsoy, YR & Spitsyn, VG 2005, Using genetic algorithm with adaptive mutation mechanism for neural networks design and training. in Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005. vol. 1, 1507882, pp. 709-714, 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005, Novosibirsk, Russian Federation, 26.6.05. https://doi.org/10.1109/KORUS.2005.1507882
Tsoy YR, Spitsyn VG. Using genetic algorithm with adaptive mutation mechanism for neural networks design and training. In Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005. Vol. 1. 2005. p. 709-714. 1507882 https://doi.org/10.1109/KORUS.2005.1507882
Tsoy, Y. R. ; Spitsyn, V. G. / Using genetic algorithm with adaptive mutation mechanism for neural networks design and training. Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005. Vol. 1 2005. pp. 709-714
@inproceedings{2a212637dc7a4a00993921e36cd52ef1,
title = "Using genetic algorithm with adaptive mutation mechanism for neural networks design and training",
abstract = "In this paper a developed evolutionary algorithm (NEvA) for simultaneous connections and weights of neural network training is described. A distinctive feature of the algorithm is flexible and effective evolutionary search and a balanced resulting neural network structure due to adaptive mutation operator. In NEvA neural network structure changes, caused by mutation operator, as well as mutation rate are defined independently for each individual. Two different problems are chosen to test the algorithm. The first one is a simple 2-bit parity problem, well known as XOR problem, and the second is a neurocontrol problem of t and 2 poles balancing. A comparison of obtained results with results of other algorithms is presented.",
author = "Tsoy, {Y. R.} and Spitsyn, {V. G.}",
year = "2005",
doi = "10.1109/KORUS.2005.1507882",
language = "English",
isbn = "0780389433",
volume = "1",
pages = "709--714",
booktitle = "Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005",

}

TY - GEN

T1 - Using genetic algorithm with adaptive mutation mechanism for neural networks design and training

AU - Tsoy, Y. R.

AU - Spitsyn, V. G.

PY - 2005

Y1 - 2005

N2 - In this paper a developed evolutionary algorithm (NEvA) for simultaneous connections and weights of neural network training is described. A distinctive feature of the algorithm is flexible and effective evolutionary search and a balanced resulting neural network structure due to adaptive mutation operator. In NEvA neural network structure changes, caused by mutation operator, as well as mutation rate are defined independently for each individual. Two different problems are chosen to test the algorithm. The first one is a simple 2-bit parity problem, well known as XOR problem, and the second is a neurocontrol problem of t and 2 poles balancing. A comparison of obtained results with results of other algorithms is presented.

AB - In this paper a developed evolutionary algorithm (NEvA) for simultaneous connections and weights of neural network training is described. A distinctive feature of the algorithm is flexible and effective evolutionary search and a balanced resulting neural network structure due to adaptive mutation operator. In NEvA neural network structure changes, caused by mutation operator, as well as mutation rate are defined independently for each individual. Two different problems are chosen to test the algorithm. The first one is a simple 2-bit parity problem, well known as XOR problem, and the second is a neurocontrol problem of t and 2 poles balancing. A comparison of obtained results with results of other algorithms is presented.

UR - http://www.scopus.com/inward/record.url?scp=33644796793&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33644796793&partnerID=8YFLogxK

U2 - 10.1109/KORUS.2005.1507882

DO - 10.1109/KORUS.2005.1507882

M3 - Conference contribution

AN - SCOPUS:33644796793

SN - 0780389433

SN - 9780780389434

VL - 1

SP - 709

EP - 714

BT - Proceedings - 9th Russian-Korean International Symposium on Science and Technology, KORUS-2005

ER -