Učební metody umělých vrstvených neuronových sítí

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27350%2F06%3A00014526" target="_blank" >RIV/61989100:27350/06:00014526 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Training Processes of Artificial Multilayer Neural Networks

Popis výsledku v původním jazyce

Artificial multilayer feed-forward neural networks are useful in many technical and scientific branches. For example, one of them is modeling and simulation. Before use any designed artificial neural network, we have to teach it for our task. The teaching process is also sometimes called training of neural network. During this teaching process some of many parameters of artificial neural network are adapted by data, which represent our task. For teaching process we can apply some special algorithms, like Back Propagation. However sometimes, during teaching process we can see situations, when is impossible stop teaching process, because error of teaching process is still "so big". The neural network can't adapt own parameters by all data, which represent our task. This paper describe some reasons of errors arise during the teaching process and also describe methods to eliminate some of these errors.

Název v anglickém jazyce

Training Processes of Artificial Multilayer Neural Networks

Popis výsledku anglicky

Artificial multilayer feed-forward neural networks are useful in many technical and scientific branches. For example, one of them is modeling and simulation. Before use any designed artificial neural network, we have to teach it for our task. The teaching process is also sometimes called training of neural network. During this teaching process some of many parameters of artificial neural network are adapted by data, which represent our task. For teaching process we can apply some special algorithms, like Back Propagation. However sometimes, during teaching process we can see situations, when is impossible stop teaching process, because error of teaching process is still "so big". The neural network can't adapt own parameters by all data, which represent our task. This paper describe some reasons of errors arise during the teaching process and also describe methods to eliminate some of these errors.

Druh

D - Stať ve sborníku

CEP obor

BC - Teorie a systémy řízení

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2006

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

4th International Workshop on Earth Science and Technology

ISBN

978-4-9902356-7-3

ISSN

—

e-ISSN

—

Počet stran výsledku

8

Strana od-do

487-494

Název nakladatele

Kyushu Univerzity

Místo vydání

Fukuoka

Místo konání akce

—

Datum konání akce

—

Typ akce podle státní příslušnosti

—

Kód UT WoS článku

—