Czasopismo | ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA | Rocznik 2014 - zeszyt 4

Global-extreme-training of specific neural networks to find laws ruling empirical data

10.15199/ELE-2014-008

nr katalogowy: 83718
10.15199/ELE-2014-008

A need to create symbolic descriptions of unknown laws governing a given set of empirical data and to model the data set features is encountered in many situations and concerns plenty of physical systems. First papers on applying particular neural networks with exp(.) and ln(.) activation functions to solve the problem under consideration ware reported in [1-8]. Recently, specific networks with reciprocal-function activation-operators have been proposed in [9], [10]. Such networks allow us to describe the data set behavior be means of rational or polynomial functions, where the latter can be regarded as a particular case of the rational one. Coefficients of the rational or polynomial functions are determined in a way of training the specific neural network. As compared to the previously presented networks, the reciprocal-function-based ones lead to simpler network structures, include a lower number of the network elements and enable more effective network learning. Despite the introduced improvements, learning the networks is still a problem, [11], [12], and we devote plenty of time to make progress in this field. Improving the training technique is the goal of our research and this is the issue to be presented in this paper. Learning problems connected with gradient calculations Neural network learning can be realized in a variety of ways [11-12]. There are methods that are good in finding global optimum and can effectively omit local extremes. Another group is made by techniques which are fast and well suited for local area operations. Ability to generalize the data set behavior outside the range of the training variables is another required feature of the learning methods. For example, so called gradient descent (GD) techniques, where gradient of an objective function is calculated, often fail in case of very large networks [12]. From our experience results that learning the networks under consideration by means of the GD[...]

Jeśli masz wykupiony/przyznany dostęp - zaloguj się. Skorzystaj z naszych propozycji zakupu!

Publikacja


e-Publikacja (format pdf) - nr 83718 "Global-extreme-training o..." licencja: Osobista Produkt cyfrowy	10.00 zł

Zeszyt


ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - e-zeszyt (pdf) 2014-4 licencja: Osobista Produkt cyfrowy	35.00 zł

Prenumerata

ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - prenumerata cyfrowa
licencja: Osobista

Produkt cyfrowy
Nowość

420.00 zł

ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - papierowa prenumerata roczna + wysyłka
licencja: Osobista

Szczegóły pakietu

Nazwa
ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - papierowa prenumerata roczna	528.00 zł brutto 488.89 zł netto 39.11 zł VAT (stawka VAT 8%)
ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - pakowanie i wysyłka	42.00 zł brutto 34.15 zł netto 7.85 zł VAT (stawka VAT 23%)

570.00 zł

ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - PAKIET prenumerata PLUS
licencja: Osobista

Szczegóły pakietu

Nazwa
ELEKTRONIKA - KONSTRUKCJE, TECHNOLOGIE, ZASTOSOWANIA - PAKIET prenumerata PLUS (Prenumerata papierowa + dostęp do portalu sigma-not.pl + e-prenumerata)	636.00 zł brutto 588.89 zł netto 47.11 zł VAT (stawka VAT 8%)

636.00 zł

Zeszyt

2014-4

Twój Koszyk

Publikacja

Zeszyt

Prenumerata

Zeszyt

Czasopisma