Dynamic Algorithms of Multilayered Neural Networks Training in Generalized Training Plant

Efimov, D. V.; Zakharenkova, T. A.; Terekhov, V. A.; Tyukin, I. Y.

Поиск по всему репозиторию:

Открыть/скачать файлы документа

93-101.pdf (533.7Kb)

Автор

Efimov, D. V.
Zakharenkova, T. A.
Terekhov, V. A.
Tyukin, I. Y.

Дата издания

1999

Издательство

BPI

Показать полное описание

Библиографическое описание

Dynamic Algorithms of Multilayered Neural Networks Training in Generalized Training Plant / D. V. Efimov [and etc.] // International Conference on Neural Networks and Artificial Intelligence ICNNAI'99 = Международная конференция по нейронным сетям и искусственному интеллекту ICNNAI'99 : Proceedings, Brest, Belarus, 12–15 October 1999 / Brest Polytechnic Institute, Department of Computers and Laboratory of Artificial Neural Networks, Belarus Special Interest Group of International Neural NetWork Society, International Neural NetWork Society, Belarusian State University of Informatics and Radioelectronics (Belarus), Belarusian Academy of Sciences, Institute of Engineering Cybemetics (Belarus), Universidad Politechnica de Valencia (Spain), Institute of Computer Information Technologies (Ukraine, Ternopil) ; ed. V. Golovko. – Brest : BPI, 1999. – P. 93–101.

Аннотация

The new modifications of multilayered neurak networks training algorithms in a generalized training plant structure are introduced. The first modification for algorithm "on error backpropagation algorithm through time" is introduced and its sufficient conditions of a training procedure stability are obtained. Other modification for speed gradient algorithm of an error backpropagation is obtained, where measurement state vector in a training procedure instead of a parametrical model of plant is used.